WHICH MEAN DO YOU MEAN? AN EXPOSITION ON MEANS

Size: px
Start display at page:

Download "WHICH MEAN DO YOU MEAN? AN EXPOSITION ON MEANS"

Transcription

1 WHICH MEAN DO YOU MEAN? AN EXPOSITION ON MEANS A Thesis Submitted to the Graduate Faculty of the Louisiaa State Uiversity ad Agricultural ad Mechaical College i partial fulfillmet of the requiremets for the degree of Master of Sciece i The Departmet of Mathematics by Mabrouck K. Faradj B.S., L.S.U., 1986 M.P.A., L.S.U., 1997 August, 004

2 Ackowledgmets This work was motivated by a upublished paper writte by Dr. Madde i 000. This thesis would ot be possible without cotributios from may people. To every oe who cotributed to this project, my deepest gratitude. It is a pleasure to give special thaks to Professor James J. Madde for helpig me complete this work. This thesis is dedicated to my wife Mariaa for sacrificig so much of her self so that I may realize my dreams. It would ot have bee doe without her support. ii

3 Table of Cotets Ackowledgmets List of Tables List of Figures Abstract ii iv v vi Chapter 1. Itroductio The Origis of the Term Mea Atique Meas Geometric Iterpretatio of the Atique Meas Atique Meas Iequality Chapter. Classical Meas The History of Classical Meas The Developmet of Classical Meas Theory Nicomachus List of Meas Pappus List of Meas A Moder Recostructio of the Classical Meas Other Meas of the Aciet Greeks Chapter 3. Biary Meas The Theory of Biary Meas Classical Meas as Biary Mea Fuctios Biary Power Meas The Logarithmic Biary Mea Represetatio of Liks betwee Biary Meas Other Biary Meas Chapter 4. -ary Meas Historical Overview The Axiomatic Theory of -ary Meas Traslatio Ivariace Property of -ary Meas Iequality Amog -ary Meas Chapter 5. Coclusio Refereces Vita iii

4 List of Tables.1 Nicomachus Meas Pappus Equatios for Meas iv

5 List of Figures 1.1 Demostratio of Atique Meas usig a circle Proof Without Words: A Truly Algebraic Iequality Biary Meas as Parts of a Trapezoid Sigle Variable Fuctio Associated With Biary Meas v

6 Abstract The objective of this thesis is to give a brief expositio o the theory of meas. I Greek mathematics, meas are itermediate values betwee two extremes, while i moder mathematics, a mea is a measure of the cetral tedecy for a set of umbers. We begi by explorig the origi of the atique meas ad list the classical meas. Next, we preset a overview of the theories of biary meas ad -ary meas. We iclude a geeral discussio o axiomatic systems for meas ad preset theorems o properties that characterize the most commo types of meas. vi

7 Chapter 1. Itroductio I the this chapter we give a brief itroductio to the origis of the arithmetic, geometric, ad harmoic meas. 1.1 The Origis of the Term Mea Accordig to "Webster s New Uiversal Dictioary", the term mea is used to refer to a quatity that is betwee the values of two or more quatities. The term mea is derived from the Frech root word mie whose origi is the Lati word medius, a term used to refer to a place, time, quatity, value, kid, or quality which occupies a middle positio.the most commo usage of the term mea is to express the average of a set of values. The term average, from the Frech word averie, is itself rich i history ad has exteded usage. The term average was used i medieval Europe to refer to a taxig system levied by a liege lord o a vassal or a peasat. The word average is derived from the Arabic awariyah, which traslates as goods damaged i shippig. I the late middle ages, average was used i Frace ad Italy to refer to fiacial loss resultig from damaged goods, where it came to specify the portio of the loss bore by each of the may people who ivested i the ship or its cargo. I this usage, it is the amout idividually paid by each of the ivestors whe a loss is divided equally amog them. The otio of a average is very useful i commerce, sciece, ad legal pursuits; thus, it is ot surprisig that several possible kids of averages have bee iveted so that a wide array of choices of a itermediate value for a give set of values is available to the user to select from. 1. Atique Meas The earliest documeted usage of a mea was i coectio with arithmetic, geometry, ad music. I the 5th cetury B.C., the Greek mathematicia Archytas gave a defiitio of the three commoly used meas of his time i his treatise o music: we have the arithmetic mea whe, of three terms, the first exceeds the secod by the same amout as the secod exceeds the third; the geometric mea whe the first is to the 1

8 secod as the secod is to the third; the harmoic mea whe the three terms are such that by what ever part of itself the first exceeds the secod, the secod exceeds the third by the same part of the third. (Thomas, 1939, p. 36) This ca be traslated to moder terms as follows. Let a ad b be two whole umbers such that a > b ad A, G, ad H are the arithmetic, geometric, ad harmoic meas of a ad b respectively. The (i) a A = A b = A = a+b, (ii) a G = G b = G = ab, a H a (iii) = H b b = H = a+b ab. The origis of the ames give to the atique meas are obscured by time. The first of these meas, ad probably the oldest, is the arithmetic mea. To the aciet Greeks, the term αριθµητικς refers to the art of coutig, ad so, fittigly, they referred to what we commoly call the average as the arithmetic mea sice it pertais to fidig a umber that is itermediate to a give pair of atural umbers. As for the ame give to the geometric mea, it appears that the Pythagorea school coied the term mea proportioal, i.e., the geometric mea, to refer to the measure of a altitude draw from the right agle to the hypoteuse of a right triagle. The measure of such a altitude is betwee the measures of the two segmets of the hypoteuse. The source of the ame give to the harmoic mea ca oly be foud i legeds. The Roma Boethius (circa 5 A.D.) tells us of a leged about Pythagoras who o passig a blacksmith shop was struck by the fact that the souds caused by the beatig of differet hammers o the avil formed a fairly musical whole. This observatio motivated Pythagoras to ivestigate the relatio betwee the legth of a vibratig strig ad the musical toe it produced. He observed that differet harmoic musical toes are produced by particular ratios of the legth of the vibratig strig to its whole. He cocluded, accordig to the leged, that the musical harmoy produced was to be foud i particular ratios of the legth of the vibratig strig. Thus to the Pythagoreas, who believed that all kowledge ca be reduced to relatios betwee umbers, musical harmoy

9 occurred because certai ratio of umbers that lie betwee two extremes are harmoic, ad thus the term harmoic mea was give to that value. Propositio 1. Suppose 0 < a b. Let A := a+b, G := ab, ad H := a+b ab. The a A b, a G b, ad a H b Proof. Sice 0 < a b, the a + b b; therefore, a+b b. Similarly, a a + b; therefore a a+b a+b. Therefore, a b. Thus, a A b. Hece, b 1 a+b 1 ab a. Therefore, a a+b b, ad a H b. If a b, a ab b ; therefore, a ab b. Hece a G b. 1.3 Geometric Iterpretatio of the Atique Meas Sice geometry is the aciet Greeks preferred veue of scietific ivestigatio, Greek mathematicias produced umerous geometric treatises that related the three atique meas to each other by usig straight edge ad compass costructio. A excellet example ca be foud i Schild (1974) ad reproduced here: Example 1.1. Suppose a ad b are two whole umbers. Let A, G, ad H be the arithmetic, geometric, ad harmoic meas respectively of a ad b. The by usig a straight edge ad compass we ca illustrate that A = a+b, G = ab, ad H = a+b ab. Draw the lie segmet LMN with LM = a ad MN = b (see figure 1.1). With LN as diameter, draw a semi circle with ceter O ad fix P o its circumferece. Draw MQ perpedicular to OP ad MP perpedicular to LN. The OP = A, MP = G, ad QP = H. To show this is true, we give the followig argumet. Sice OP is the radius of the circle whose diameter is LN, the OP = 1 (a + b) = A, ad sice (MP) = (LM)(MN) = ab, the MP = ab = G. Let α = POM. Observe that QMP = α ad POM is similar to PMQ; thus, PQ PM = PM PO. Therefore, PQ = (PM) PO = ab a+b = ab a+b = H. I figure 1.1, observe what happes if (a + b) remais fixed, i.e., segmet LN is fixed, ad M is allowed to move. As M moves toward N, both G ad H decrease. As M moves towards O, both G ad H icrease. If M coicides with O, i.e., a = b, the A = G = H. This may have bee the motivatio for ivestigatig the iequality betwee the three meas. 3

10 P Q α α L O M N a b FIGURE 1.1. Demostratio of Atique Meas usig a circle. 1.4 Atique Meas Iequality I this sectio we will preset several proofs of the iequality: H G A (1.1) Of the umerous useful iequalities i mathematics, the arithmetic-geometric mea iequality occupies a special positio, ot oly from a historical stadpoit, but also o accout of its frequet usage i differet mathematical proofs. We will give a more i-depth discussio about this iequality i Chapters 3 ad 4. At this poit, it suffices to say that there have bee umerous proofs give for the above iequality over the ceturies. We begi our discussio by presetig a iformal argumet of the iequality. Referrig back to Figure 1.1, we ote that siα 1. From OPM, we have siα = MP OP, ad from PQM, we have siα = QP MP. Therefore MP = OPsiα G = Asiα. Hece G A, (1.) ad QP = MPsiα H = Gsiα. Hece H G. (1.3) From 1. ad 1.3, we get 1.1. However, sice the above argumet uses trigoometry, it does ot reflect the spirit of the aciet proofs for this iequality. I Figure 1., we preset a illustratio that captures the fudametal 4

11 character of this iequality i mathematics, which may have motivated the aciet mathematicias to establish proofs of the arithmetic-geometric mea iequality (Gallat 1977). The iequality as illustrated by Figure 1. requires oly rudimetary kowledge of geometry to prove. Now we give a more moder algebraic proof for the geometric-arithmetic mea iequality. ab a b FIGURE 1.. Proof Without Words: A Truly Algebraic Iequality. Theorem 1. For ay oegative umbers a ad b, ab a+b, with equality holdig if ad oly if a = b. Proof. Let a = c ad b = d. The a+b ab becomes c +d cd, or equivaletly, c +d cd 0. This is equivalet to c cd + d 0 which is i tur equivalet to (c d) 0. Sice the square of ay real umber is oegative, we see that the iequality stated i the theorem is ideed true. Equality holds if ad oly if c d = 0, that is c = d, or equivaletly, if ad oly if, a = b. We use the result from theorem 1 to establish a iequality betwee the harmoic ad geometric meas of ay two oegative umbers. Corollary 1.. For ay oegative umbers a ad b, ab a+b ab, with equality holdig if ad oly if a = b. Proof. Sice ab a+b, the ab (a + b). Therefore, ab (a + b) ab, ad ab a+b ab. 5

12 From theorem 1 ad corollary 1., we have H G A. (1.4) 6

13 Chapter. Classical Meas I this chapter we will explore the origis of the theory of biary meas. The chapter icludes two lists of the classical biary meas as give by Greek mathematicias. The followig list gives the ames of Greek mathematicia ad the approximate dates of their work o meas. It is helpful to the uderstadig of the historical developmet of the theory meas i the aciet Greek world (Smith 1951). Thales, 600 B.C. Pythagoras, 540 B.C. Archytas, 400 B.C. Plato, 380 B.C. Eudoxus, 370 B.C. Eudemus, 335 B.C. Euclid, 300 B.C. Archimedes, 30 B.C. Hero, 50 A.D. Nicomachus, 100 A.D. Theo, 15 A.D. Porphyrius, 75 A.D. Pappus, 300 A.D. Iamblichus, 35 A.D. Proclus, 460 A.D. Boethius, 510 A.D..1 The History of Classical Meas I this sectio we will give a brief discussio o what motivated Greek mathematicias to study ad develop a doctrie for meas by presetig the ratioale give by promiet Greek mathematicias who touched o the history of the theory of meas i their work ad the opiios of Greek mathematics scholars o this matter. Accordig to Gow (193), by Plato s time umbers were grouped ito two geeral categories. First, as sigle umbers categorized by their attributes such as odd, eve, triagular, perfect, excessive, defective, amicable etc. Secod, umbers were viewed as groups comprised of umbers that are either i series or proportios. The aciet Greeks viewed meas as a special case of proportios (Allema 1877, Thomas 1939, Gow 193). Smith (1951) writes, " Early [Greek] writers spoke of a arithmetic proportio, meaig b a = d c as i,3,4,5, ad of geometric proportio, meaig a : b = c : d as i, 4, 5, 10, ad a harmoic proportio, meaig 1 b a 1 = d 1 1 c as i 1, 1 3, 4 1; 1 5." I his commets o paradigms of aciet Greek mathematics, Allema (1877) says, "whe two quatities were compared [i Greek mathematics], the basis for the compariso seems to be either how much the oe is greater tha the other, i.e., a arithmetic 7

14 ratio, or how may times is the oe cotaied i the other, i.e., their geometrical ratio." Allema (1877) claims that this type of compariso of ratios would aturally lead to the theory of meas because for ay three positive magitudes, be it lies or umbers, a, b, ad c, if a b = b c, the three magitudes are i arithmetical proportio, but if a : b :: b : c, they are i geometrical proportio. Allema s claim seems to be supported by the work of Nicomachus i "Itroductio to Arithmetic". I this work, Nicomachus bega his discourse o meas by givig the defiitio that distiguished a ratio from a proportio. He referred to the latter as the compositio of two ratios. He the stated that whe oe term appears o both sides of a proportio, as i b a = b c, the proportio is kow as a cotiued proportio. The proportio is called disjuct whe the middle terms are differet. The highest term i a cotiued proportio is called the cosequet, the least is called the atecedet, ad the middle term is the mea, µεστητες, which is medius whe traslated ito Lati ad from which the word mea is derived (Gow 193). As we have oted above, Greek mathematics viewed meas as a special proportio ivolvig three magitudes; therefore, it is appropriate that we begi our review of the history of developmet of meas by metioig that Proclus attributed to Thales the begiig of the doctrie of proportios (Allema 1877). Thales established the theorem that equiagular triagles have proportioal sides (Allema 1877). I "Itroductio to Arithmetic", Nicomachus writes, "the kowledge of proportios is particularly importat for the study of aciet mathematicias." This ca be take to mea that the doctrie of proportios played a importat role i the developmet of Greek mathematics. Maziarz (1968) commets o the atural developmet of the theory of proportioals i Greek mathematics by sayig, "If a poit is a uit i a positio, the a lie is made of poits. Cosequetly, the ratio of two give segmets is merely the ratio of the umber of poits i each. Moreover, because ay magitude ivolves a ratio betwee the umber of uits it cotais ad the uit itself, ad, thus, the compariso of two magitudes implies either or 4 ratios." By poits, Maziarz seems to imply the tick marks that would be made if the segmets were divided ito may small equal uits. 8

15 From the historical perspective, the aciet sources of Greek mathematics history that we have refereced do ot metio whe the arithmetic mea was first developed. However, they offer various explaatios as to whe the geometric ad harmoic meas were first itroduced. Allema (1877) states that aciet sources (Iamblichus, Nicomachus, Proclus) poit to Eudoxus as the oe who established the harmoic mea ad to Pythagoras as the oe who established the otio of a mea proportioal betwee two give lies. It is iterestig to ote that some facets of the theory of meas appear i various aciet Greek texts. Some of these were iteded as mathematics treatises, such as the collectio of books that costitute Euclid s work kow as the "Elemets", but others did ot have a apparet mathematical purpose. Oe such example, oted by Maziarz (1968), ca be foud i passages of "Timaeus" kow as "The Costructio the world-soul." I this sectio of the book, Plato attempts to costruct the arithmetical cotiuum usig two geometric progressios 1,,4,8 ad 1,3,9,7; the fillig i the itervals betwee these umbers with the arithmetic ad harmoic meas. By successive duplicatio of the two progressios ad fillig i with the appropriate combiatio of arithmetic ad harmoic meas, all umbers ca be geerated, but ot i their atural order. Aother example ca be foud i Aristotle s "Metaphysics". I this work, Aristotle describes Plato s otio of distributive justice as, " The just i this sese is a mea betwee two extremes that are disproportioate, sice the proportioate is a mea, ad the just is proportioate. This kid of proportio is termed by mathematicias geometrical proportio." From the above examples, oe gets the sese that to the aciet Greeks, the theory of meas ad proportios may ot have bee just a mere mathematical cocept sice some aspects of the theory of meas was also reflected i their literature, philosophy, ad religio.. The Developmet of Classical Meas Theory It appears that the classical meas were developed over a log period of time by the gradual additio of seve more meas to the first three (Heath 1963). I all his work, Euclid oly uses the three atique meas (Allema 1887, Gow 193). However, by first cetury A.D., we kow that 9

16 Greek mathematicias referred to te meas. All the sources reviewed (Allema 1887, Bema 1910, Heath 191, Gow 193, Thomas 1939, Smith 1951) suggest that Greek mathematicias geerated these meas by cosiderig three quatities a, b, ad c, such that a > b > c. They assumed b to be the mea ad formed three positive differeces with the a, b, ad c: (a b), (b c), ad (a c). The they formed a proportio by equatig a ratio of two of these differeces to a ratio of two of the origial magitudes, a, b, ad c. For example, b is the harmoic mea of a ad c whe a b b c = a c. Nicomachus i "Itroductio to Arithmetic" (Gow 193) goes o to say: "Pythagoras, Plato, ad Aristotle kew oly six kids of [cotiued] proportios: the arithmetic, geometric, ad harmoic meas, ad their subcotraries, which have o ames. Later writers added four more." Greek mathematicias referred to certai classical meas as cotrary ad subcotrary meas because these meas were see to be i a cotrary (opposite) order from the arithmetic mea whe compared to the geometric or harmoic meas (Oxford Eglish Dictioary 004). I his work "I Nicomachus" (Heath 191), Iamblichus says, "the first three [atique meas] oly were kow to Pythagoras, the secod three were iveted by Eudoxus." The remaiig four, Iamblichus attributed to the later Pythagoreas. He adds that all te were treated i the Euclidea maer by Pappus. Gow (193) states that the umber of cotiued proportios was raised to te ad kept at that umber because the umber te was held by the aciet Greek mathematicias to be the most perfect umber. He adds, "how else ca we explai the fact that the golde mea, which Nicomachus calls the most perfect ad embracig of all proportios, was left out from the list of meas." All these testimoies poit to the coclusio that the theory of meas i Greek mathematics was well established by the First Cetury. Our mai complete source for aciet Greek mathematics theory of meas is Boethius commetary o the works of Pappus ad Nicomachus. I this work, 10

17 Boethius credits Nicomachus ad Pappus as the mai Greek mathematicias who dealt with meas from a theoretical perspective (Smith 1951)..3 Nicomachus List of Meas The earliest kow treatmet of classical meas as a idepedet body of kowledge was give by Nicomachus i "Itroductio to Arithmetic" (Allema 1887, Heath 191, Gow 193, Thomas 1939, Smith 1951). Allema, Gow, Heath, ad Thomas cocluded (seemigly idepedet of each other) that Nicomachus proceeded to develop his list as follows: He bega his list by commetig o the cotiued arithmetical proportio a b = b c. This suggests that a b : b c :: a : a, which allows us to make a coectio to other meas. Gow (193) remarks, "I a cotiued geometric proportio, a : b :: b : c, he otices that a b : b c :: a : b. Fially, the three magitudes, a, b, c, are i harmoic proportio if a b : b c :: a : c." A similar approach was used by Archytas (as cited by Porphyrius i his commetary o Ptolemy s "Harmoics") whe discussig the three atique meas i terms of three magitudes i cotiued arithmetic, geometric, ad harmoic proportios (Thomas 1939). Gow (193) also poits out that Nicomachus failed to metio that the arithmetic, geometric, ad harmoic meas of two umbers are i geometric proportio: a+b : ab : a+b ab. I Thomas traslatio of Nicomachus "Itroductio to Arithmetic" (Thomas 1939), Nicomachus itroduces the seve other meas usig the same treatmet as the oe metioed above. (The reader may wish to refer to Table.3 for a compact summary of the followig.) The fourth mea, which is also called the subcotrary by reaso of its beig reciprocal ad atithetical to the harmoic, comes about whe of the three terms the greatest bears the same ratio to the least as the differece of the lesser terms bears to the differece of the greater, as i the case of 3, 5; 6 (Thomas, 1939, p. 119). Nicomachus itroduces the fifth mea as the subcotrary mea to the geometric mea, The fifth [mea] exists whe of the three terms, the middle bears to the least the same ratio as their differece bears to the differece betwee the greatest ad the middle 11

18 terms, as i the case of, 4; 5, for 4 is double, the middle term is double the least, ad is double 1, that is the differece of the least terms is double the differece of the greatest. What makes it subcotrary to the geometric mea is this property, that i the case of the geometric mea the middle term bears to the lesser the same ratio as the excess of the greater term over the middle bears to that of the middle term over the lesser, while i the case of this mea a cotrary relatio holds (Thomas, 1939, p. 11). Nicomachus itroduces the sixth mea as, The sixth mea comes about whe of the three terms the greatest bears the same ratio to the middle as the excess of the middle term over the least bears to the excess of the greatest term over the middle as i the case of 1, 4; 6, for i each case the ratio is sesquialter [3 : ]. No doubt, it is called subcotrary to the geometric mea because the ratios are reversed, as i the case of the fifth mea (Thomas, 1939, p. 11). Nicomachus itroduces the last 4 meas by sayig, By playig about with the terms ad their differeces certai me discovered four other meas which do ot fid a place i the writigs of the aciets, but which evertheless ca be treated briefly i some fashio, although they are superfluous refiemets, i order ot to appear igorat. The first of these, or the seveth i the complete list, exists whe the greatest term bears the same relatio to the least as their differece bears to the differece of the lesser terms, as i the case of 6, 8; 9, for the ratio of each is see by compoudig the terms to be the sesquialter. The eighth mea, or the secod of these, comes about whe the greatest term bears to the least the same ratio as the differece of the extreme bears to the differece of the greater terms, as i the case of 6, 7; 9, for here the two ratios are the sesquialter. The ith mea i the complete series, ad the third i the umber of those more recetly discovered, comes about whe there are three terms ad the middle bears to the least the same ratio as the differece betwee the extremes bears to the differece betwee the least terms, as 4, 6; 7. Fially, the teth i the 1

19 complete series, ad the fourth i the list set out by the moders, is see whe i three terms the middle term bears to the least the same ratio as the differece betwee the extremes bears to the differece of the greater terms, as i the case of 3, 5; 8, for the ratio i each couple is the super-bi-partiet [5 : 3] (Thomas, 1939, p. 11). TABLE.1. Nicomachus Meas Mea Proportio Numbers Exhibitig the Mea Arithmetic a b : b c :: a : a, 4, 6 Geometric a b : b c :: a : b 4,, 1 Harmoic a b : b c :: a : c 6, 3, Cot. Harmoic b c : a b :: a : c 3, 5, 6 Cot. Geometric b c : a b :: b : c, 4, 5 Subco. Geometric b c : a b :: a : b 1,4,6 Seveth a c : b c :: a : c 6, 8, 9 Eighth a c : a b :: a : c 6, 7, 9 Nith a c : b c :: b : c 4, 6, 7 Teth a c : a b :: b : c 3, 5, 8 (Thomas 1939).4 Pappus List of Meas Pappus used a differet approach tha Nicomachus whe presetig his list of meas (Heath 191, Thomas 1939). Both Heath ad Thomas state that the meas o Pappus list are similar to those preseted by Nicomachus, but i a differet order after the sixth mea. Meas umber 8, 9, ad 10 i Nicomachus list are respectively umbers 9, 10, ad 7 o Pappus list. Moreover, Pappus omits mea umber 7 o Nicomachus list ad gives as umber 8 a additioal mea equivalet to the proportio c : b :: c a : c b. Therefore, the two lists combied give five additioal meas to the first six. I Thomas traslatio (1939) of Pappus work kow as "Collectios III", Pappus itroduces his discussio o meas as a respose to a questio posed by a uiformed geometer. He 13

20 demostrates his aswer by the costructio of the three meas i a semicircle (see figure 1.1). Pappus shows, i a series of propositios, that give three terms α, β, ad γ i geometrical progressio (Heath 191 uses "i geometric proportio"), it is possible to form from them three other terms a, b, ad c which are itegral liear combiatio of α, β, ad γ such that b is oe of the classical meas. The solutios to Pappus s equatios are show i Table.. The liear (Heath 191, Thomas 1939) TABLE.. Pappus Equatios for Meas Mea a, b, c Numbers exhibitig the mea Arithmetic a = α + 3β + γ 6, 4, b = α + β + γ c = β + γ Geometric a = α + β + γ 4,, 1 b = β + γ c = γ Harmoic a = α + 3β + γ 6, 3, b = β + γ c = β + γ Subcotrary a = α + 3β + γ 6, 5, b = α + β + γ c = β + γ Fifth a = α + 3β + γ 5, 4, b = α + β + γ c = β + γ Sixth a = α + 3β + γ 6, 4, 1 b = α + β + γ c = α + β γ Seveth a = α + β + γ 3,, 1 b = β + γ c = γ Eighth a = α + 3β + γ 6, 4, 3 b = α + β + γ c = β + γ Nith a = α + β + γ 4, 3, b = α + β + γ c = β + γ Teth a = α + β + γ 3,, 1 b = β + γ c = γ equatios show i Table. are moder equivalets of the literal traslatio of the Greek versio of Pappus. For example (Thomas 1939), i the case of the geometric mea metioed i Table., the literal traslatio of Pappus words would be, "To form a take α oce, β twice, ad γ oce; ad to form b we have to take β oce ad γ oce; ad to form c we take γ oce." Notice also that the examples give by Pappus for the proportios formed by his equatios sometimes differ 14

21 from those give by Nicomachus. For example for the fourth mea, Nicomachus gave 3, 5, ad 6 as a example for a solutio, while Pappus gave, 5, ad 6 as a solutio. Pappus expositio o meas by usig equatios may be better uderstood from the perspective that proportios were used i those days to solve equatios. Usig Proclus commetary o Euclid as a referece, Klei (1966) states, " Greek mathematics usage of proportios ca be compared to the moder sese of costructio of a equatio, ad a equatio may be viewed as a solutio of a proportio. This may be due to the uderstadig of ratios, proportios, ad harmoy o the basis of a commo mathematical property." Bema (1910) claims that the mathematicias of Alexadria uderstood equatios of secod degree mostly i the form of proportios. If we express Pappus method i moder terms, Pappus is parmeterizig meas by quadratics ad, equivaletly, givig quadratic polyomials to illustrate the relatio amog terms i the various meas. For example, to calculate the harmoic mea, usig three quatities i geometric progressio is equivalet to usig α = 1, β = x, ad γ = x ; thus, give a = + 3x + x, b = x + x, ad c = x + x, we have ac a+c = (+3x+x )(x+x ) = x(x+1)(x +3x+) = x(x +3x+) +3x+x +x+x (x+1) x+1 = x(x + ) = x + x = b..5 A Moder Recostructio of the Classical Meas I this sectio, we will use a similar approach to the oe used by Nicomachus to geerate the classical meas by cosiderig three positive quatities a, b, ad c such that a > b > c, ad we wish to make b the mea of a ad c. We will form three positive differeces with these quatities: (a b), (b c), ad (a c). The we will form a proportio by equatig a ratio of two of these differeces to a ratio of two of the origial quatities (ot ecessarily distict). For example, if we set the ratio a b b c equal to the ratio a b, the result is b = ac, which represets the geometric mea. If you look at all the possible ways of doig this, several of them are automatically ruled out by the assumed iequality of a, b, ad c. The oes that are ot (ecessarily) ruled out are the eleve meas summarized below (Madde 000, Heath 1963): 1. (a b) (b c) = a a = b b = c c (a+c), we have the arithmetic mea b =. 15

22 . (a b) (b c) = b c = a b, we have the geometric mea b = ac (a b) (b c) = a c ; we have the harmoic mea b = 1 a +. 1 c (a b) (b c) = c a ; we have the cotra-harmoic mea b = a +c a+c. (a b) (b c) = c b ; we have the first cotra-geometric mea b = a c+ a ac+5c. (a b) (b c) = b a ; we have the secod cotra-geometric mea b = c a+ 5a ac+c. (b c) (a c) = a c; b = ac c a. This mea is o Nicomachus list but ot Pappus list. (b c) (a c) = c b ; b = c+ 4ac 3c. (a b) (a c) = c a ; b = a ac+c a. (a b) (a c) = b a ; b = a c a. This mea is o Pappus list but ot Nicomachus list. (a b) (a c) = c b ; b = a c. Note that some of these meas are ot very robust defiitios of meas. For example, if oe uses the 11 th mea o our list to fid the mea of 5 ad 4, the M(5,4) = 1, which is ot betwee 5 ad 4. Note also that usig the 5 th mea o our list to fid the mea of 1 ad, we obtai the celebrated golde umber Φ = However, as we will show i the ext sectio, the above list does ot exhaust all the meas kow to the aciet Greek world..6 Other Meas of the Aciet Greeks I this sectio, we poit out that Greek mathematicias cotiued to develop ew meas which were ever icluded amog the classical meas. Nicomachus referred to a special mea obtaied by the divisio of a segmet ito what he called "the most perfect proportios". This mea, which we will call b, ca be expressed by the divisio of a segmet of magitude a ito two parts: A greater part, b, ad a lesser part, a b, i such a fashio that the ratio of a to b is equal to the ratio 16

23 of b to a b. Hece the proportio: a : b = b : (a b) This, i tur, leads to the quadratic equatio b + ab a = 0, The positive root of which is b = 1 a 5 1. A special solutio of this equatio is whe a = 1, b is the celebrated umber Φ. The mathematicias of Alexadria referred to other quatities as meas. For example Hero s mea. Defiitio.3. Suppose a ad c are positive umbers. The Hero s mea is b = a + ac + c. 3 To check that Hero s mea of ay two positive values is always betwee these two values, let 0 < a < c. The by the arithmetic-geometric mea iequality, a < ac < a+c < c. Thus a = 3a 3 < a + ac 3 < a + ac + c 3 < c + ac 3 < 3c 3 = c. Hero s mea is used i calculatig the volume of a pyramidal frustum (a prismatoid figure formed by choppig off the top of a pyramid), where a ad c are the bottom ad top areas respectively of the pyramidal frustum. The cetroidal mea is aother example of a mea produced by aciet Greek mathematics which was ot icluded i the list of classical meas. This mea was developed by Archimedes for his work o cetroids. Defiitio.4. Let a ad c be two atural umbers. The the cetroidal mea of a ad c is b = (a + ac + c ). 3(a + c) 17

24 We shall demostrate that the cetroidal mea of two positive values is always betwee these two values. Let 0 < a < c. Usig the iequality from propositio 1, we have a a+c ac c; therefore, a 3 + (a + c ) 3(a + c) ac 3(a + c) + (a + c ) 3(a + c) c 3 + (a + c ) 3(a + c). Now, a < c a + ac < a + c a a +c a+c a < 3 a + (a +c ). A similar argumet ca be used to show c c 3 + (a +c ) 3(a+c). Hece, a (a +c ) 3(a+c) c. 3(a+c) 18

25 Chapter 3. Biary Meas I this chapter we will give a cotemporary defiitio for biary meas ad preset o overview of the developmet of the theory of biary meas. The chapter icludes a expositio of the most commo types of biary meas. The chapter cocludes with a summary o iequalities amog biary meas. 3.1 The Theory of Biary Meas Based upo the sources reviewed, aciet Greek mathematics treatmet of meas teded to be limited to fidig the mea of two magitudes, be it lie segmets, areas, or volumes. Berlighoff (00) claims that this limited view o meas i Greek mathematics may have stemmed from their iterest i geometry, where meas of magitudes of segmets, areas, ad volumes are itermediate value betwee the two extremes. Therefore, fidig the mea of more tha two such magitudes was a problem that was ot ecoutered because such a mea would ot represet itermediate value betwee two extremes. I moder times, this outlook has chaged. The arithmetic mea, geometric mea, ad the harmoic mea came to be viewed as specific cases of a geeral fuctio ot of just two variables but also of -variables. I this chapter we will limit our discussio to biary meas ad postpoe our dealig with -ary meas to the ext chapter. Hutigto (197) cites work published by R. Schimmack i 1909 which treats meas as a cotiuous fuctio that satisfies give restrictios. Dodd (1933) credits B. de Fietti s 1931 work with formulatig specific criteria that a mea fuctio must satisfy. Although i both cases, the fuctio referred to is a variable fuctio, a similar view may be exteded to two-variable mea fuctios. Borwei (1987) lists postulates for a mea fuctio of two variables, f (a,b), similar to the restrictios cited by Hutigto ad Dodd for a mea fuctio of -variables. We will use Borwei s (1987) defiitio ad criteria for biary meas to develop a defiitio for a geeralized biary mea. Next we will subject the classical meas to the criteria we have 19

26 developed for biary meas to coclude whether or ot these meas ca be cosidered meas i our refied sese. We will also itroduce the moder otio of power meas, ad show the various iequalities that relate biary meas to each other. We will coclude the chapter with examples of other fuctios that geerate biary meas. We begi by itroducig the term isotoe (Borwei 1987), which we will subsequetly use to idetify a specific property for fuctios of two variables. Defiitio 3.5. Let f : R + R + R + be a fuctio. f is isotoe if for each a R + ad b R +, f (a,x) ad f (x,b) are mootoe icreasig fuctios of x. To demostrate that f (a,b) is isotoe, we fix oe variable, say a, ad show that f (a,x) is mootoe icreasig as a fuctio of x. The we appeal to the same argumet for b. Defiitio 3.6. A biary mea fuctio, f, is a positive real valued fuctio, f (a,b), of two strictly positive real variables a ad b that satisfies the followig postulates: CR f is a cotiuous ad real valued fuctio. IS IN f is a isotoe. f is iteral, i.e., mi(a,b) f (a,b) max(a,b). DI f is diagoal, i.e., f (a,b) = a or f (a,b) = b if ad oly if a = b. HO f is homogeeous, i.e., f (λa,λb) = λ f (a,b), where λ 0. SY f is symmetric, i.e., f (a,b) = f (b,a). Remark: Note that HO permits us to write M(a,b) = am(1, b a ), a useful result utilized i proofs of may theorems o meas. 3. Classical Meas as Biary Mea Fuctios We will revisit our eleve classical meas to explore which of these satisfy the biary mea postulates listed i defiitio 3.6. For the followig argumets, we will assume that a ad b are positive real umbers such that a < b ad M is the mea. Propositio. If M is equal to A, G, or H, the M is a biary mea fuctio. 0

27 Proof. Suppose 0 < a b ad M(a,b) is A, G, or H. Clearly, M satisfies CR. To show M satisfies IS, fix a ad let 0 < x 1 < x. Now we show M(a,x 1 ) < M(a,x ). If M = A or M = H, the M(a,x 1 ) < M(a,x ) because a+x 1 < a+x ad ax 1 a+x 1 < ax a+x by properties of additio ad multiplicatio of positive umbers. If M = G, we have x 1 < x ad ax 1 < ax. Thus, ax1 < ax, sice the square root fuctio is mootoe icreasig. Thus M(a,x 1 ) < M(a,x ). A similar argumet ca be used to show M(x,b) is mootoe icreasig. M satisfies IN by propositio 1. M satisfies DI. This ca be checked by substitutig b = a i the defiitio of A, G, ad H. M satisfies HO by the distributive property of multiplicatio over additio. M satisfies SY by the commutative properties of additio ad multiplicatio. The remaiig eight classical meas are ot biary mea fuctios accordig to defiitio 3.6. To substatiate this claim, we take each i turs. 1. The cotra-harmoic mea, M = a +b a+b, fails to satisfy IS. For example, M(6,) = 5 = M(6,3).. The cotra-geometric mea, M = (a b + a ab + 5b )/ fails to satisfy SY. For example, M(1,) = = O the other had, M(,1) = = The subcotra-geometric mea, M = (b a + 5a ab + b )/, fails to satisfy SY. For a example, M(1,4) = = O the other had, M(4,1) = = M = ab b a. M fails to satisfy SY. For example, M(1,) = 4 4 = 0. O the other had, M(,1) = 4 1 = M = b+ 4ab 3b also fails SY. 6. M = a ab+b a. Clearly, this mea fails to satisfy SY. 7. M = a b a. M fails to satisfy SY. 1

28 8. M = b a. Clearly, M fails to satisfy SY. Therefore, of the eleve classical meas, oly the atique meas are cosidered biary mea fuctios accordig to defiitio 3.6for a mea. 3.3 Biary Power Meas Aother represetatio for biary mea fuctios is kow as power meas. The a root-mea-square (also kow as the Euclidea mea), R(a,b) = +b, may have bee the first example of this ew class of meas (Li 1974). Defiitio 3.7. Suppose r > 0, a > 0, ad b > 0. The the r th power mea of a ad b, deoted M r (a,b), is ( ar +b r ) 1 r. Theorem. Let a > 0, b > 0, ad r 0. The fuctio M r (a,b) = ( ar +b r ) 1 r is a biary mea fuctio. Proof. M r (a,b) statisfies CR. Clearly M r (a,b) is cotiuous for r > 0, a > 0 ad b > 0 sice it is a compositio of cotiuous fuctios. M r (a,b) satisfies IS. Fix a. Let x 1 ad x be ay positive umbers such that 0 < x 1 < x. If r > 0, the x r 1 < xr ad ar +x r 1 < ar +x r ( ) a r +x r r ( ) 1 a < r +x r r. If r < 0, the x r 1 > x r ad ar +x r 1 > ar +x r. Therefore,. Therefore, ( a r +x r 1 M r (a,b) satisfies IN. Suppose a < b. If r > 0, the we have a < ( ar +b r ) 1 r < b sice ) r < ( a r +x r a r < ar +b r < b r. Similarly, if r < 0, we have a r > ar +b r > b r ad a < ( ar +b r ) 1 r < b. M r (a,b) satisfiesdi. Suppose ( ar +b r ) 1 r = a. The we have ar +b r = a r which implies b = a. Similarly, if b = a, the ( ar +a r ) 1 r = a. Therefore, M r (a,b) = a if ad oly if b = a. M r (a,b) satisfies HO. Fix λ > 0. The M r (λa,λb) = ( (λa)r +(λb) r ) 1 r = λ( ar +b r ) 1 r. M r (a,b) satisfies SY, sice M r (a,b) = ( ar +b r ) 1 r = ( br +a r ) 1 r = M r (b,a). ) r. Note that the arithmetic mea, the harmoic mea, ad the root-mea-square are power mea fuctios by direct substitutio i M r = ( ar +b r ) 1 r with the appropriate value for r: 1. r = 1, the M 1 (a,b) yields the arithmetic mea A = a+b.

29 . r = 1, the M 1 (a,b) yields the harmoic mea H = ab a+b. 3. r =, the M (a,b) yields the root-mea-square R = a +b. We ow show that the geometric mea is a limit of power mea fuctios. Theorem 3. lim r 0 M r (a,b) = ab. { Proof. Observe that lim( ar +b r r 0 ) 1 r = lim exp ( 1 r 0 r )l( ar +b r Applyig L Hopital rule, lim ( = lim ) ( ) a r la+b r lb r 0 a r +b r = ab. ( l(a r +b r )/ r 0 r ) = lim r 0 = la+lb. Therefore, exp } { = exp ) ) ( ddr ((l(a r +b r ))/) 1 { (lim r 0 ( 1 r )l( ar +b r ) lim ( 1 r 0 r )l( ar +b r ) } }. = exp { } la+lb Defiitio 3.8. Let a ad b be ay positive umbers. The M 0 (a,b) := ab. With the developmet of this represetatio for meas, ways had to be foud to compare these meas to each other ad to the already established oes. This led to the to the establishmet of some of the most well-kow iequalities i mathematics. Theorem 4. If a, b, ad r are positive umbers, the M 0 (a,b) M r (a,b). With equality holdig if ad oly if a = b. Proof. Note that a = b ( ) ab = a r +b r 1 r. Suppose a < b, the M 0 (a,b) = ab ad ( ) M r (a,b) = a r +b r 1 r. Observe that ab = (ab) 1. The (ab) r = (a r b r ) 1 ad [ ( ) a r +b r 1 ] r r = ar +b r. By the arithmetic-geometric mea iequality, (a r b r ) 1 < ar +b r. Therefore, M 0 (a,b) < M r (a,b). Theorem 5. If a, b, r, ad s are positive umbers such that r < s, the M r (a,b) < M s (a,b). The proof we preset is a modified versio of the proof give i Schaumberger (1988) for -ary power meas. 3

30 Proof. Let x > 0 ad f (x) = rx s + (s r) sx r. We ote that f (x) has a absolute miimum oly at x = 1 (sice f (x) = rsx s 1 rsx r 1 = 0 oly at x = 1 ad f (1) = rs(s r) > 0). Observe that f (1) = 0; therefore, f (x) = rx s + (s r) sx r 0. Hece, rx s + (s r) sx r, (3.5) with equality holdig if ad oly if x = 1. Let T = ( a r +b r ) 1 r. Put x 1 = a T ad x = b T. By substitutig for x 1 ad x i equatio 3.5 successively for x ad addig, we obtai [ ( r at ) s ( + bt ) s ] [ ( + s r s at ) r ( + bt ) r ] [ ] [. Hece, r a s +b s T s + s r s a r +b r T ]. r But [ ] ( ) T r = ar +b r. Therefore, a s +b s T s. Hece as +b s T s, ad this implies as +b s a r +b r s r, which ( ) leads to a s +b s 1 ( ) s a r +b r 1 r. Therefore, M r (a,b) M s (a,b). 3.4 The Logarithmic Biary Mea The logarithmic mea is ecoutered i various applicatios such as i ivestigatio of heat trasfer, fluid mechaics (Li 1974), ad the distributio of electrical charge o a coductor (Stolarsky 1975). Defiitio 3.9. Let a > 0 ad b > 0. The L(a,b) = a a b la lb if a b if a = b Theorem 6. L is a biary mea. Proof. First we prove that L(a,b) satisfies CR. Clearly, L(a,b) is cotiuous o (0, ) (0, ) except maybe o the lie a = b. To show that L(a,b) is cotiuous whe a = b, we ote first that ) ) = 1 by L Hopital rule. Thus lim = a; therefore, by substitutig x y for u, lim u 1 ( u 1 lu we have lim (x,y) (a,a) ( y( x y 1) l x y a > 0 ad let x (0, ). Let g(x) := L(a,x) = ) (y,u) (a,1) y( u 1 lu = a. So, lim L(x,y) = a. Now we show L(a,b) satisfies IS. Fix (x,y) (a,a) x a. We must show that g(x) is mootoe lx la = x a l a x 4

31 icreasig. It suffices to show g (x) > 0, except possibly at fiitely may poits. Whe x a, g (x) = l a x (1 a x ) (l a x ; so, it suffices to show l a x (1 a x ) ) (l a x > 0, except at fiitely may poits. Let ) h(u) = lu 1 + u. We eed to show h(u) > 0, except at fiitely may poits. By examiig h, we see that h is decreasig o (0,1) ad icreasig o (1, ). Sice h(1) = 0, we have proved what is eeded. We show that L(a,b) satisfies IN, i.e., mi(a,b) L(a,b) max(a,b). Let f (x) = l(x). The for ay 0 < a < b, by the Mea Value Theorem, there exists a t i [a,b] such that f (t) = lb la b a. Therefore, 1 t = la lb a b. Hece t = la lb a b ad a t b. That L(a,b) satisfies DI is evidet from the defiitio of L(a,b). We show that L(a,b) satisfies HO. Let λ > 0. The L(λa,λb) = λa λb logλa lλb = λ(a b) We show that L(a,b) satisfies SY. L(a,b) = a b la+lλ la lλ) = λ( la lb ) = λl(a,b). a b la lb = (b a) (lb la) = The followig theorem establishes a iequality betwee L, A ad G. b a lb la = L(b,a). Theorem 7. If a > 0 ad b > 0 such that a b,the G(a,b) < L(a,b) < A(a,b). The followig proof was give by Carlso (197) Proof. If t > 0, the iequality of the arithmetic ad geometric mea implies that t +t(a + b) + ( a+b ) > t +t(a + b) + ab > t + t(ab) 1 + ab. Thus R dt < R (t+ 0 a+b ) 0 fractios, we fid ab < dt (t+a)(t+b) < R a b la lb < a+b 0 a+b < a b 1 lim dt (t+ ab). Evaluatig the middle itegral by the method of partial [l(t + b) l(t + R a)]r 0 < 1. This implies ab Based upo the results obtaied above, we have the followig iequality that relates the harmoic mea, geometric mea, logarithmic mea, arithmetic mea, ad the root-mea-square. Corollary Let a > 0 ad b > 0 such that a > b, the H G L A R. It is iterestig to ote that the logarithmic mea does ot quite led itself to a atural geeralizatio to variables (Pitteger 1985). This mea fails a particular axiom (amely the associativity axiom) for -ary meas. However, due to the use of this -ary mea i various applicatios such as i defiig average temperatures ad aalysis of idex umbers i 5

32 ecoomics, a theoretical framework for the geeralizatio of the logarithmic mea of variables has bee recetly developed. We refer the iterested reader to the work of Pitteger (1985) for more iformatio o -ary logarithmic meas. 3.5 Represetatio of Liks betwee Biary Meas Eves (003) gives a excellet geometric lik betwee various biary meas usig a trapezoid. Let a > b > 0. Suppose a trapezoid has parallel sides a ad b as show i figure 3. The various a b ab C R T N G H FIGURE 3.3. Biary Meas as Parts of a Trapezoid meas ca be raked i size relative to each other as the legths of vertical segmets. The segmet whose legth is: The harmoic mea, H, passes through the itersectio of the diagoals. The geometric mea, G, divides the trapezoid ito two similar trapezoids. The Heroia mea, N, is oe third of the way from the arithmetic mea to the geometric mea. The arithmetic mea, A, bisects the sides of the trapezoid. The cetroidal mea, T, passes through the cetroid of the trapezoid. The root-mea-square, R, bisects the area of the trapezoid. 6

33 The cotra-harmoic mea, C, is as far to the right of the arithmetic mea as the harmoic mea is to the left of it. 3.6 Other Biary Meas Iterest i geeratig differet biary meas fuctios cotiued to grow ito the late 0 th cetury as other fuctios of two variables were foud that satisfy give criteria for a desired mea fuctio. Borwei (1987) defied a class of biary mea fuctios, M p (a,b), that is derived from a mea fuctio, M(a,b), that satisfies the postulates give i Sectio 3.1. This class of biary meas is determied by the formula M p (a,b) := M(ap,b p ) M(a p 1,b p 1 ). where p R. We refer the reader to Borwei (1987) for the proof that M p (a,b) satisfies the postulates give i Sectio 3.1. Example of such biary meas iclude (Borwei 1987): Lehmer meas. Let a,b > 0 ad p R. The Lehmer meas, L p, is defied as L p (a,b) = ap + b p a p 1 + b p 1. Observe that L 1 = A ad L 1 = G. Gii meas. Let a,b > 0 ad r s. The Gii mea, G (s,r) (a,b), is defied as ( a s + b s ) ( 1 s r) G (s,r) (a,b) = a r + b r Stolarsky s Meas. Let a,b > 0 ad p 0,1. The the Stolarsky s Mea, S p (a,b), is defied as ( a p + b p S p (a,b) = p(a b) ) ( ) 1 p 1 7

34 Observe that S 0 (a,b) = lim S p (a,b) = b a p 0 lb la, which is the logarithmic mea. Ad S 1 (a,b) = lim S p (a,b) = e 1 (a a b b ) a b 1, which is also kow as the idetric mea. p 1 We refer the iterested reader to Borwei (1987) for more iformatio o the meas listed above. Mays (1983) ivestigated coditios uder which biary meas ca be associated with a sigle variable fuctio. I this work, Mays developed a idea preseted by Moskovitz (1933). Mays also poited out some errors cotaied i Moskovitz (1933). Give a fuctio f from (0, ) ito R, Mays (1983) defies M f (a,b) to be the X-itercept of the lie coectig (a, f (a)) ad (b, f (b). See Figure 3.4 (reproduced from Mays (1983)). Clearly M f (a,b) satisfies IN ad SY. We fid a formula for M f (a,b) by calculatig the slope of the lie through (a, f (a)) ad (b, f (b) i two ways: Solvig for M f (a,b), we get: f (a) a M f (a,b) = f (b) b M f (a,b). M f (a,b) = a f (b) + b f (a) f (a) + f (b). (3.6) Theorem 8. M f = M g g = k f for some k > 0 Proof. If g = k f, k cacels i the right had side of M k f (a,b) = ak f (b)+bk f (a) k f (a)+k f (b) = M g = M f. If g k f, pick a, b, ad k so that g(a) = k f (a) but g(b) k f (b). The if M f (a,b) = M g (a,b), we ] ] have =. Therefore, a f (a)(k f (b) g(b)) = b f (a)(k f (b) g(b)). [ a f (b)+b f (a) f (a)+ f (b) [ ag(b)+bk f (a) k f (a)+g(b) Sice k f (b) g(b) 0 ad f (a) 0, the a = b, a cotradictio. Corollary If M = M f, the there exists f such that M = M f ad f = 1. Proof. Let f = f (x) f (1). Corollary 3.11 allows us to assume, without loss of geerality, whe associatig M f with a give fuctio f that f (1) = 1. Defiitio 3.1. Let f be a fuctio i oe variable. f is multiplicative if the domai of f is closed uder multiplicatio ad f (xy) = f (x) f (y) for every x, y i the domai of f. 8

35 FIGURE 3.4. Sigle Variable Fuctio Associated With Biary Meas Lemma Suppose f : (0, ) = R ad f (1) = 1. The f is multiplicative if ad oly if f (a) f (λb) = f (λa) f (b) (3.7) for all a, b, ad λ (0, ). Proof. Note that the coditio 3.7 implies f (λb) = f (λ) f (b) for all λ, b (0, ). Coversely, if f is multiplicative, the f (a) f (λb) = f (a) f (λ) f (b) = f (λa) f (b). Theorem 9. Suppose f (1) = 1. The M f is homogeeous if ad oly if f is multiplicative. Proof. By equatio 3.6, a, b, λ > 0 λm f (a,b) = M f (λa,λb) λ ( a f (b)+b f (a) f (a)+ f (b) ) = a f (b) f (λa) + a f (b) f (λb) + b f (a) f (λa) + b f (a) f (λb) = λa f (λb)+λb f (λa) f (λa)+ f (λb) af(a)f(λb) + b f (a) f (λa) + a f (b) f (λb) + b f (b) f (λa) a f (b) f (λa) + b f (a) f (λb) = a f (a) f (λb) + b f (b) f (λa) (a b) f (b) f (λa) = (a b) f (a) f (λb) f (b) f (λa) = f (λb) f is multiplicative. 9

36 We ow explore some ideas that are motivated by Mays (1983). Let M be a fuctio of two variables (ot ecessarily a mea). We defie F M (x,y) := M(x,y) y x M(x,y). Similarly, if F is a fuctio of two variables, we defie M F (x,y) := xf(x,y) + y 1 + F(x,y). Note that if M is a mea, F is the kid of ratio that was cosidered by the Greeks i developig the classical meas. Theorem 10. F MF = F ad M FM = M as fuctios o {(x,y) x y}. If we choose M ad F such that M F = M ad F M = F, the: 1. M is homogeeous if ad oly if F is "projective", i.e., F(x,y) = F(λx,λy).. M is itermediate if ad oly if F is positive. 3. M is symmetric if ad oly if F(x,y)F(y,x) = 1. Proof. We show F MF = F. F MF = M F y x M F = xf+y 1+F y x xf+y 1+F = xf+y y yf x+xf xf y = (x y)f (x y) = F. We show M FM = M. M FM = xf M y M+y x( x M )+y 1+F M = 1+( M y x M ) = x(m y)+xy My x M+M y = (x y)m (x y) = M. Now we show that M is homogeeous if ad oly if F(x,y) = F(λx,λy). Suppose M is homogeeous. The F(λx,λy) = M(λx,λy) λy λx M(λx,λy) = λm(x,y) λy λx λm(x,y) = λ(m(x,y) y) F(x,y) = F(λx,λy). M(λx,λy) = λxf(λx,λy)+λy 1+F(λx,λy) λ(x M(x,y)) = M(x,y) y ( x+f(x,y)+y = λ 1+F(x,y) x M(x,y) = F M(x,y). Now suppose ) = λm(x, y). We ow show M is itermediate if ad oly if F > 0. x(1 + F) > xf + y > y ad M y x M > 0 m is betwee x ad y. To show M is symmetric if ad oly if F(x, y)f(y, x) = 1, let F = F(x, y) ad F = F(y, x). Observe that M F is symmetric xf+y 1+F = yf+x xf + y + xff + yf = yf + x + yff + Fx 1+F x + yff = y + xff (x y) = (x y)ff FF = 1. 30

37 Mays addressed the problem of whe a give biary mea m ca be expressed as M f for some f (x). We ow use F ad M to give a more geeral solutio to the problem tha Mays. Propositio 3. Suppose M is ay fuctio of two variables x ad y such that x y. The M = M f if ad oly if F M (x,y) = f (y) f (x). Proof. Suppose M = M f. The m = xf(x,y)+y 1+F(x,y) = x f (y)+y f (x) f (x)+ f (y). Therefore, x f (x)f(x,y) + y f (x) + x f (y)f(x,y) + y f (y) = y f (x)f(x,y) + y f (x) + x f (y)f(x,y) + x f (y), ad x f (x)f(x,y) + y f (y) = y f (x)f(x,y) + x f (y). Hece, (x y) f (x)f(x,y) = (x y) f (y), which implies F(x,y) = f (y) f (y) f (x). Coversely, we ca show if F(x,y) = f (x), the M f = xf(x,y)+y 1+F(x,y) = x f (y)+y f (x) f (x)+ f (y) let M = M f, we have proved what is eeded. by usig a similar argumet to the oe give above. Therefore, if we 31

38 Chapter 4. -ary Meas The chapter begis with a brief discussio of the historical developmet of -ary meas. Next we preset a overview of postulates of -ary meas, startig with postulates for the arithmetic mea ad follow their evolutio ito postulates for geeralized meas. We discuss the traslatio ivariace property for -ary meas. We preset examples of various types of -ary meas. We coclude the chapter with a brief discussio of the theory of iequalities amog -ary meas. 4.1 Historical Overview Oe of the earliest kow refereces cocerig the arithmetic mea of several umbers is give by Iamblichus i a treatise o what we call ow umber theory. I this work, "The Theology of Arithmetic," Iamblichus outlies a example ivolvig fidig the arithmetic mea of the umbers 1 to 9: I the first place, we must set out i a row the sequece of umbers from the moad up to ie: 1,, 3, 4, 5, 6, 7, 8, 9. The we must add up the amout of all of them together, ad sice the row cotais ie terms, we must look for the ith part of the total to see if it is already aturally preset amog the umbers i the row; ad we will fid that the property of beig [oe] ith [of the sum] oly belogs to the [arithmetic] mea itself (Heath, 191, p.8). As we have metioed i Chapter 1, i the middle ages the term average referred to the equal apportiomet of a loss or expese icurred by a ship (or its cargo), i which case the idividual compesatio made by the owers (or isurers) of a ship or its cargo is i proportio to the value of their respective iterests. This otio of average represets the most documeted usage of the arithmetic mea durig that period. I the 17 th cetury, astroomers were makig several observatios of specific cosmic evets for cofirmatio purposes. They were faced with the problem of combiig observatios to come up with a sigle value that best represeted the true value of the quatity beig measured. Hald 3

39 (1998) states that for a log period of time the usual practice [by astroomers] for estimatig the true value was to select the best amog several observatios of the same object, the best beig defied by such criteria as the occurrece of good observatioal coditios, the exertio of special care, ad so o. Gradually, however, it became commo practice to use the arithmetical mea as a estimate of the true value. No theoretical foudatio for this practice seems to have existed before the works of Simpso ad Lagrage i the mid 18 th cetury. The problem of fidig the best estimate of a ukow parameter from a set of direct observatios of that parameter may be very difficult. It depeds o the distributio of the parameter, ad (as Gauss showed) oly whe the distributio is ormal is the arithmetic mea i every case the best estimate. (The precise sese of "best" is itself a complex problem that we shall avoid addressig). Therefore, the arithmetic mea is ot always the best choice for averagig a set of observatioal data. Steves (1955) states that i choosig a method of averagig physical magitudes, oe fudametal issue to be cosidered is the atural method of combiig them. Where magitudes are aturally combied by takig sums, the arithmetic mea is meaigful ad may be useful. However, where positive magitudes are aturally combied by takig products, the geometric average may be the most appropriate to use. To establish a geeral framework for the presetatio of the various meas ad the postulates o -ary meas, we itroduce the followig: Covetio Let X R, where R is the set of real umbers ad i N. We cosider a sequece of fuctios f i : X i R. For coveiece, we sometimes write f (a,b,...,c) lettig f represet the appropriate f i. We call X the domai of f. It will be useful for us at this poit to defie the aalogues for the most commo meas ecoutered: Defiitio The arithmetic mea, A = {A i } i N is: A(x 1,x,...,x ) = A = x 1 + x x. 33

40 The geometric mea, G is: G(x 1,x...,x ) = G = (x 1 x...x ) 1. The harmoic mea, H is: H(x 1,x...,x ) = H = 1 x 1 + x x 1 = 1 A(x 1 1,x 1,...,x 1 ). The root-mea-square, R, of the sequece is: R(x 1,x...,x ) = R = ( 1 (x 1 + x x )) 1. Remark Note that A ad R are defied for X = R. While G is defied oly for X = R 0 := {x R x 0}, ad H is defied for X = R >0 := {x R x 0}. The first three meas show above are clearly similar to the classical meas, while the root-mea-square is a obvious geeralizatio of the biary root-mea-square. The arithmetic mea ad the root-mea-square are widely used i mechaics (as i the defiitios of the ceter of gravity ad radius of gyratio), ad i the moder theory of statistics. The root-mea-square of the differeces of some variable from its arithmetic mea is the stadard deviatio (x i x) i=1. The geometric mea is used i the costructio of idex umbers i ecoomics. The harmoic mea is little used, except i special ivestigatios (Hutigto 197). A atural geeralizatio of these meas is referred to as power meas, sometimes kow as Cauchy meas (Bulle 198): 34

41 Defiitio Let r R \ {0}. The r th power mea is: M r = [ 1 i=1 x r i ] 1r. Remark Note that the domai of M r always cotais {x R x > 0}, but i some cases it is larger. Uless specified otherwise, we will take M r to refer to a sequece of fuctios with domai {x R x > 0}. Thus M r satisfies axiom PO (see below). I the cases r = -1, 1, ad, M r is the harmoic mea, the arithmetic mea ad the root-mea-square respectively. Although r is defied to be ozero, the followig theorem establishes that the geometric mea is a limitig case of M r as r teds to 0. Theorem 11. Let x 1,...,x be positive real umbers. The lim r 0 M (x 1,x,x 3,...,x ) = (x 1 x x 3...x ) 1 r. The followig proof is from Burrows (1986). Proof. Let y(r) = 1 x r i, where r 0. The y (r) = 1 x r i lx i. By the Mea Value Theorem i=1 i=1 y(r) = y(0) + ry (θ), where 0 < θ < r. Hece y(r) = 1 + ry (θ). Now, 1 r ly(r) = 1 r l(1 + ry (θ) = y (θ) + o(r) = 1 r (ry (θ) + +o(r ) 1 = lim r 0 r ly(r) = y (0) = 1 i=1 lx i. Sice y (r) is cotiuous. Therefore, takig the atilogarithm of this last result we get lim y(r) = (x 1x...x ) 1 r. r 0 35

42 4. The Axiomatic Theory of -ary Meas Begiig aroud 1900, several authors took up the problem of fidig ad aalyzig axiomatic characteristics for various -ary meas. The followig postulates appear i the various postulate systems we have reviewed, ad, thus, we give them special labels for coveiece. We demad the equatios to be true whe the terms are defied, i.e. whe all argumets belog to X. PO The domai of f is X = {x R x > 0} ad f i (x 1,x...,x i ) > 0. SY f is symmetric, i.e., it is idepedet of the order i which the quatities x 1,x,...,x X, are take, i.e., f (x 1,x...,x i,x j,...,x ) = f (x 1,x...,x j,x i,...,x ). DI f is diagoal, i.e., f (a,a,a,...,a) = a. IN f is iteral, i.e., a f (x 1,...,x ) b if a x i b for all i. HO f is homogeeous, i.e., for all k, f (kx 1,kx,kx 3,kx 4,...,kx ) = k M(x 1,x,...,x ), where x i X. OD f is odd, i.e., f ( x 1,..., x ) = f (x 1...x ) (Note that this a special case of HO). TR f is traslatio ivariat, i.e, f (k + x 1,...,k + x ) = k + f (x 1...x ) for ay k. AS f is "associative" i the sese that f (x 1,x...,x ) = f ( f i,..., f } {{ } i,x i+1,...x ), i times where f i = f (x 1,...x i ). AS f (x 1,x,x 3,x 4,...,x ) = M(m,m,x 3,x 4,...,x ), where m = f (x 1,x ). The earliest approach to the theory of meas by usig the postulatio method is Schimmack (1909 p. 18). He gave a set of axioms that completely characterize the arithmetic mea of positive umbers. Specifically, he proved the followig theorem (Schimmack 1909): Theorem 1. Let f be a sequece of fuctios such that f satisfies T R, OD, SY, ad AS. The f is the arithmetic mea. We refer the reader Schimmack (1909) for a elegat proof of the above theorem. Beetle (1915) established the complete idepedece of Schimmack s postulates. Beetle (1915) states, "The 36

43 otio of complete idepedece is much more restrictive tha the requiremet of idepedece. The requiremet for the latter is that o oe property is a logical cosequece of ay of the others. However, these properties are ot ecessarily devoid of iterrelatios. For example, it may well be that o-possessio of oe property implies possessio of aother. Complete idepedece implies either ay oe of them, or its egative, is a logical cosequece of ay combiatio formed by the others ad their egatives." To show the complete idepedece of Schimmack s four postulates, Beetle (1915) proved the existece of 4 types of systems, f, each defied ad real valued for all real values of its argumets, i which at least oe system possess ay give combiatio of the properties but does ot possess the remaiig properties. Gratta-Guiess (000) refers to Hutigto as oe of the major America postulatioists whose mai mathematical iterest was developig axiomatic systems for various mathematical cocepts ad establishig their cosistecy, idepedece, completeess, ad equivalece. Hutigto (197) exteded Schimmack s work by cosiderig fuctios f that satisfy the geeral postulates give below. He established the idepedece of these postulates i a maer similar Beetle s (1915) method i establishig the idepedece of Schimmack s postulates. Defiitio Hutigto s geeral postulates are: PO, HO, DI, SY, ad AS. We call f a Hutigto mea if it satisfies Hutigto s geeral postulates. Hutigto cocered himself with the arithmetic mea, geometric mea, the harmoic mea ad the root-mea-square. We will geeralize some of his results to power meas, which we defied earlier. Theorem 13. Suppose M r is the r th power mea, where r is a ozero real umber. The M r is a Hutigto mea. Proof. M r satisfies PO, sice each x i is positive. M r satisfies SY by the commutative law of additio. 37

44 [ ] k M r satisfies HO. Let k > 0. The M r (kx 1,kx,...,kx ) = r x r 1 +kr x r +...+kr x r 1 r = [ x r ] k 1 +x r xr r = km r. M r satisfies DI, sice M r (a,a,...,a) = a. M r satisfies AS. Suppose M r (x 1,x ) = m.the m = xp 1 +xp. Therefore, [ x r ] M r (x 1,x,x 3,x 4,...,x ) = 1 +x r 1 [ x r +...+xr r 1 +x r ] + xr 1 +xr 1 +x = x r r = [ m+m+x r ] x r 1 r = M r. Theorem 14. G is a Hutigto mea. Proof. G = (x 1 x...x ) 1. The G satisfies PO, sice the product ad powers of positive umbers are positive. G satisfies SY, sice multiplicatio is commutative. To show G satisfies HO, let k > 0. The ((kx 1 )(kx )...(kx )) 1 = (k (x 1 x x 3...x )) 1 = k(x 1 x x 3...x ) 1 = kg. G satisfies DI, sice G(a,a,...,a) = a. G satisfies AS. Suppose G = (x 1 x ) 1 = m x 1 x = m. The G = (x 1 x x 3...x ) 1 = G = (x 1 x x 3...x ) 1 = (m x 3...x ) 1 As we have metioed earlier, Hutigto (197) established several other properties that completely characterize each of the four meas, A, G, H, ad R. The followig theorem summarizes some of the results that Hutigto preseted. Theorem 15. Let f be a Hutigto mea. The: a) f = A if ad oly if f (1 x 1,1 x,...,1 x ) = 1 f (x 1,x,...,x ). b) f = H if ad oly if x 1 f ( x 1 1, x x 1,..., x x 1 ) = f (x 1,x,...,x ) f (x 1,x,...,x ) 1. 38

45 c) f = G if ad oly if f ( 1 x 1, 1 x,..., 1 x ) = 1 f (x 1,x,...,x ). d) f = R if ad oly if f ((1 x 1) 1,(1 x ) 1,...,(1 x ) 1 ) = (1 ( f (x1,x,...,x )) ) 1. The followig theorem geeralizes parts (a), (b), ad (d) of theorem 15. Theorem 16. Let f be a Hutigto mea. Suppose r 0. The f = M r if ad oly if f ((1 x r 1 ) 1 r,(1 x r ) 1 r,...,(1 x r ) 1 r ) = (1 ( f (x1,x,...,x )) r ) 1 r. (4.8) for all 0 < x i < 1. Proof. First we prove that if f = M r, the M r satisfies equatio 4.8. We have ( ) 1 M r ((1 x r 1 ) 1 r,(1 x r ) 1 r,...,(1 x r ) 1 r r ) = = ( 1 xr 1 +xr +...+xr ) 1 r = (1 (M r (x 1,x,...,x ) r ) 1 r. ((1 x 1 ) 1 r) r +((1 x r ) 1 r ) r +...+((1 x r ) 1 r ) r To prove the coverse, assume equatio 4.8. First, we show f (a,b) = M r (a,b). ( ) f (a,b) = (a r + b r ) 1 a b r f, (a r + b r ) 1 r (a r + b r ) 1 r b r (by HO) = (a r + b r ) 1 r f ((1,(1 (a r + b r ) 1 r (a r + b r ) 1 r (by algebraic idetities) ( { ( )} r ) 1 r = (a r + b r ) 1 b a r 1 f, (a r + b r ) 1 r (a r + b r ) 1 r (by 4.8)3.6. a r = ((a r + b r ) f (b,a) r ) 1 r (by HO & simple maipulatio). ) 39

46 Thus f (a,b) r = (a r + b r ) f (b,a) r. So f (a,b) r = (a r + b r ), ad f (a,b) r = ar +b r. Therefore, ( a r + b r ) 1 r f (a,b) =. Now we prove that f = M r for ay umber of argumets. Suppose that x 1,x,...x are give. We claim that for k = 1,..., there is a q such that x r 1 + xr xr k (k 1)qr > 0. To prove the claim, ote that x r 1 + xr xr k (k 1)qr > 0 for k =,..., if ad oly if x r 1 + xr xr k > (k 1)qr if ad oly if xr 1 +xr +...+xr k k 1 > q r, for k =,..., if ad oly if ( x r 1 +x r +...+xr k k 1 ( x r 1 +x r +...+xr k k 1 ) 1 r > q whe r > 0 ) 1 r < q whe r < 0 for k =,...,. So it is oly ecessary to pick q satisfyig fiitely may iequalities. The claim is thus proved. Now, M r (q,(x r 1 + xr qr ) 1 r ) = Mr (x 1,x ). Let Z k = (x r 1 + xr xr k (k 1)qr ) 1 r. Our choice of q, esures that Z k is the r th root of a positive umber. Also, from f (a,b) = M r (a,b), we have f (x 1,x ) = f (q,z ), ad i geeral f (Z k,x k+1 ) = f (q,z k+1 ), for k = 1,..., 1. So, from AS, we have 40

47 f (x 1,x,...x ) = f (q,z,x 3,...x ) = f (q,q,z 3,x 4,...x ) =... (4.9) = f (q,q,q,...,q,z ) Now, put a = M r (x 1,...,x ). The a = f (a,a,...,a) by DI = f (q,q,...,q,(a r ( 1)q r ) 1 r )) by 4.9 ad xi = a = f (x 1,x,...,x ) by 4.9 Now we give a proof part (c) of theorem 15. Proof. First, we show that G( 1 x 1, 1 x,..., 1 x ) = 1 G(x 1,x,...,x ). (4.10) We have G(x 1,x,x 3,...,x ) = (x 1 x x 3...x ) 1. Therefore, G( 1 x 1, 1 x,..., 1 x ) = ( 1 x 1 1 x... 1 x ) 1 1 = ( ) 1 x 1 x...x 1 = (x 1 x...x ) 1 1 = G(x 1,x,x 3,...,x ). 41

48 Now we prove the coverse. Suppose f ( 1 x 1, 1 x,..., 1 x ) = 1 f (x 1,x,...,x ). (4.11) Let a ad b be positive umbers. Usig equatio 4.11 for =, we have f ( 1 a, 1 b ) = 1 f (a,b) Multiplyig by ab ad usig HO, we get f (b,a) = f ( ab a, ab ) = ab b f (a,b) But by SY ad HO, the left had side is f (a,b) = f (b,a). Thus f (a,b) = ab ad f (a,b) = (ab) 1. (4.1) f (x 1,x,x 3,...,x ) = f (1,x 1 x,x 3,...,x ) by 4.9 ad AS = f (1,1,x 1 x x 3,x 4,...,x ) by 4.9 ad AS =... by 4.9 ad AS = f (1,1,...,1,x 1 x...x ). Set a = (x 1 x x 3...x ) 1 ad let each of the x i = a, we see that f (a,a,...,a) = f (1,1,...,1,a ) = a, by DI. Hece f (x 1,x,x 3,...,x ) = (x 1 x x 3...x ) 1. 4

49 Aother milestoe work o the properties of geeralized -ary meas that is cotemporary with Hutigto (197) is Nagumo (199). Nagumo proved the followig theorem: Theorem 17. Suppose M is a sequece of fuctios of real umbers satisfyig the postulates: SY, DI, AS, ad IN i additio, the property: For ay x 1 < x, we demad x 1 < M(x 1,x ) < x (a stregtheig of IN). The M is of the form: M(x 1,x,...,x ) = ϕ 1 ( i=1 ϕ(x i ) ), where ϕ(x) is a cotiuous mootoe icreasig fuctio with iverse ϕ 1. Dodd(1934) proved the complete idepedece of the postulates for meas of the type give by Nagumo (199). 4.3 Traslatio Ivariace Property of -ary Meas Hoeh & Nive (1985) proved that certai meas other tha A, while failig to satisfy T R, satisfy "traslatioal iequalities". Their theorem is : Theorem 18. Let A be the arithmetic mea, G be the geometric mea, H be the harmoic mea, ad R be the root-mea-square. Let a 1,a...,a,x be positive umbers, where the a i s are ot all equal. The: (i) A(x + a 1,x + a,...,x + a ) = x + A(a 1,a,...,a ). (ii) G(x + a 1,x + a,...,x + a ) > x + G(a 1,a,...,a ). (iii) H(x + a 1,x + a,...,x + a ) > x + H(a 1,a,...,a ). (iv) R(x + a 1,x + a,...,x + a ) < x + R(a 1,a,...,a ). First Hoeh & Nive proved lemma 4.0, i which they also proved part (i) of theorem 18 ad used the results from the lemma to prove the parts (ii) of theorem

50 We will preset the proof from Hoeh & Nive (1985) for lemma 4.0 ad part (ii) of theorem 18. After this, we shall state ad prove a theorem about M r that geeralizes parts (iii) ad (iv) of theorem 18. Proof. I part (ii) of theorem 18, we must show: G(x + a 1,x + a,...,x + a ) > x + G(a 1,a,...,a ). First apply the mea value theorem to the differetiable fuctio G(x) o the iterval [0,c]. Thus, G(a 1 +c,a +c,...,a +c) G(a 1,a,...,a ) c 0 = G (a 1 + θ,a + θ,...,a + θ) for 0 < θ < c. But G > 1 by part (b) of lemma 4.0. Therefore, G(a 1 +c,a +c,...,a +c) G(a 1,a,...,a ) c 0 > 1 ad G(a 1 + c,a + c,...,a + c) G(a 1,a,...,a ) > c + G (a 1,a,...,a ). Lemma 4.0. Let a 1,a...,a,x be positive real umbers, where the a i s are ot all equal. The: (a) da dx = 1, ad (b) dg dx > 1. Proof. Part (a): First, we will establish that A(x) = A(x + a 1,x + a,...,x + a ) = x + A(a 1,a,...,a ). A(x + a 1,x + a,...,x + a ) = x + a 1 + x + a x + a = x + a 1 + a a = x + a 1 + a a, = x + A(a 1,a,...,a ). Therefore, differetiatig A(x) with respect to the variable x we have da dx = d dx A(x + a 1,x + a,...,x + a ) = x + A(a 1,a,...,a ) = 1. 44

51 Part (b): For the geometric mea, ote that the derivative of the product i=1 (a i + x) is the ( 1) elemetary symmetric polyomial of the (a i + x), deoted here by S 1. For example, i the case = 3 the polyomial S would be S = (a 1 + x)(a + x) + (a 1 + x)(a 3 + x) + (a + x)(a 3 + x). I geeral, S 1 = ( j i (a j + x). i=1 Let G(x) = G(x + a 1,x + a,...,x + a ); therefore G = i=1 (a i + x). Thus, d dx G = G 1 dg dx = S 1. To prove that dg dx > 1, it suffices to show S 1 > G 1. This is othig more tha the arithmetic-geometric mea iequality applied to the terms of S 1 = [ j i (a j + x) ]. The geometric mea of these terms is the th root of their product, i=1 which is [(a 1 + x)(a + x)...(a + x)] 1, or G ( 1), because each (a j + x) appears i exactly 1 terms of S 1. Theorem 19. Suppose M r (x 1,x,...,x ) = ( xr 1 +xr +...+xr ) 1 r, where r 0. Also, suppose a 1,a,...,a are positive real umbers (ot all the same). The M r (x + a 1,x + a,...,x + a ) > M r (a 1,a,...,a ) if r < 1, ad M r (x + a 1,x + a,...,x + a ) < M r (a 1,a,...,a ) if r > 1. To prove the above theorem 19, we will use lemma 4.1 (the proof of lemma 4.1 will follow below). Proof. Whe r < 1. By lemma 4.1, d dx m r(x) > 1. The M r is a mootoe icreasig fuctio, ad together with a 1,a...,a,x beig positive real umbers, we have: M r (x + a 1,x + a,...,x + a ) > M r (a 1,a,...,a ). Whe r > 1. Similarly, by lemma 4.1, d dx m r (x) < 1. The M r is mootoe a decreasig fuctio, ad together with a 1,a...,a,x beig positive real umbers, we have: 45

52 M r (x + a 1,x + a,...,x + a ) < M r (a 1,a,...,a ) if r > 1. Lemma 4.1. Let a 1,a...,a,x be positive real umbers, where the a i s are ot all equal, ad r 0. Suppose m r (x) = M r (x + a 1,x + a,...,x + a ): The d dx m r(x) > 1 whe r < 1, ad d dx m r(x) < 1 whe r > 1. Proof. Suppose 0 < r < s. The M r (y 1,y...,y ) = ( yr 1 +yr +...+yr ) 1 r, ad usig the iequality, we have ( yr 1 + yr yr ) 1 y s r 1 + ys ys ) 1 s. Hece, m r (x) > m r 1 (x). This implies that (m r (x)) r 1 > 1 < ((x + a 1) r 1 + (x + a ) r (x + a ) r 1 ) r > 1 as r < 1 Now (m r (x)) r 1 ((x + a 1) r + (x + a ) r (x + a ) r ). So r(m r (x)) r 1 d dx m r(x) = r ((x + a 1) r 1 + (x + a ) r (x + a ) r 1 ); Thus, d dx m r(x) = 1 (x + a 1 ) r 1 + (x + a ) r (x + a ) r 1 (m r (x)) r 1 > < 1 as r > 1 r < 1. Based upo the theorem 18 ad lemma 4.0, Hoeh & Nive (1985) wet o to ote that for every positive value x, (G(x) A(x)) < 0 although (G(x) A(x)) is a icreasig fuctio, ad (R(x) A(x)) > 0 although (R(x) A(x)) is a decreasig fuctio. This observatio motivated 46

53 Hoeh & Nive (1985) to look at the limits of these fuctios as x teds to ifiity. Hoeh & Nive (1985) oted that the equatio: lim [F(x + a 1,x + a,...,x + a ) A(x + a 1,x + a,...,x + a ) = 0 x holds with ay oe of A, G, H, or R i place of F. Hoeh & Nive (1985) established the result by the followig theorem: Theorem 0. lim x [F(x + a 1,x + a,...,x + a ) x] = A(a 1,a,...,a ) holds with ay oe of A, G, H, or R i place of F. Proof. We begi with the case of F(x) = A(x). We have lim [F(x + a 1,x + a,...,x + a ) x] = lim [A(x + a 1,x + a,...,x + a ) x]. x x But x = A(x + a 1,x + a,...,x + a ) A(a 1,a,...,a ) by lemma 4.1. Therefore, lim [A(x + a 1,x + a,...,x + a ) x] = A(a 1,a,...,a ). x I the other three cases, it suffices to show F = H(x) ad F = R(x) ad usig the iequality H < G < R for the case F = G. To prove F = H, we use the result from calculus: c m x m + c m 1 x m c 0 lim x k m x m + c m 1 x m 1 = c m. (4.13) k 0 k m assumig k 0 0, we write H = H(x + a 1,x + a,...,x + a ) i the form H = G S 1 where G is the product give above ad S 1 is as above. Thus we write G xs 1 lim (H x) = lim. (4.14) x x S 1 Now S 1 is a polyomial i x of degree 1. The coefficiet of x 1 is, ad the coefficiet of x is ( 1)(a 1 + a a ). The coefficiet of x i G is 1, ad x 1 is (a 1 + a a ). Hece, i G xs 1, the terms of degree cacel, ad by applyig 4.13 to 4.14, we have lim x (H x) = lim G xs 1 x S 1, ad we get lim x (H x) = (a 1+a +...+a ) = A. 47

54 To prove the case F = R, we write lim x [R(x + a 1,x + a,...,x + a ) x] = lim(r x) = lim R x R+x. Expadig R x as a quadratic polyomial i x, we otice that the x terms cacel, so that R x = (a 1+a +...+a ) + (a 1 +a +...+a ). It follows that R x R+x (a 1+a +...+a) = + (a 1 +a +...+a ) x R x+1 we have divided the umerator ad deomiator by x. As x we ote that lim R x calculatio R x = 1 [( a 1 x + 1) + ( a x + 1) ( a x + 1) ]. Hece we have lim [R(x + a 1,x + a,...,x + a ) x] = lim(r x) = lim R x x R+x = (a 1+a +...+a) = A., where = 1 from the 4.4 Iequality Amog -ary Meas Oe of the most promiet property of -ary meas, at least from the theoretical stad poit, is of course the iequalities betwee the various meas. This topic has attracted the attetio of may mathematicia as evidet by the richess of the research i that area. We begi the sectio with the arithmetic-geometric mea iequality for -ary meas. Cauchy appears to have bee the first to state the theorem i its most geeral form. The theorem is listed i Cauchy s Cours d Aalyse (p ), which appeared i 181. We will preset two differet proofs for this iequality. The we give some commetary o this iequality. Next we exted this iequality to iclude the harmoic mea, ad we close the sectio with a theorem o iequalities betwee power meas. Theorem 1. Let x 1,x,x 3...,x be o-egative real umbers. The 1 (x 1 + x x ) (x 1 x...x ) 1, ad equality occurs if ad oly if all the x i s are equal. The first proof we preset is by Chrystal (1916). We chose this proof because it was used i a umber of college freshma algebra textbooks aroud the tur of the 0th cetury ad represets a elemetary approach to provig the iequality. Proof. Let a,b,...k be a sequece of o-egative real umbers. Cosider their geometric mea, (ab...k) 1. If a,b,...k are ot all equal, replace their greatest ad least of them, say a ad k, by a+k. 48

55 The ( a + k ) > ak, ad the result has bee to icrease the geometric mea of the sequece of umbers, while the arithmetic mea of the quatities, a+k a+k,b,c,..., is clearly the same as the arithmetic mea of a,b,c,...k. If the ew set of quatities are ot equal, replace the greatest ad the least as before, ad so o. By repeatig the process sufficietly ofte, we ca make all the quatities as early equal as we please, ad the the geometric mea of the sequece of umbers becomes equal to the arithmetic mea. But, sice the arithmetic mea remaied ualtered throughout, ad the geometric mea has bee icreased at each step, it follows that the first geometric mea, amely (abc...k) 1, is less tha the arithmetic mea, a+b+...+k. The ext proof we have selected is the oe give by Thacker (1851). We chose this proof because it is cocurret with Chrystal s proof, but it shows a approach that is more mathematically balaced yet still straight forward. Proof. If x > 0 ad is a iteger. The usig the biomial theorem, we ote the followig: (1 + x ) = 1 + x x + 1 ( 1)(1 ) x ,ad (4.15) 3 1 (1 + x 1 ) 1 = 1 + x x + 1 ( 1 1 )(1 3 Note that for ay positive iteger, we have the followig iequalities 1 ) x (4.16) 1 1 > 1 1 1, 1 > 1 1,... >... 49

56 Hece every term ivolvig i the series expasio i 4.15 is greater tha the correspodig term i the series expasio i 4.16, ad sice the terms of series i 4.15 ad 4.16 are positive, we have (1 + x ) > (1 + x 1 ) 1. (4.17) Now, let there be positive quatities a 1,a,...,a arraged i order of magitude with a 1 beig the least. The ( a1 + a + a a ) ( = a a 1+a +...+a a 1 a 1 ) ( ) 1 > a 1+a1 +a +...+a a 1 1 a 1 ( 1) by 4.17 Note that a 1 ( 1+a1 +a +...+a a 1 a 1 ( 1) ) 1 ( = a +a a a ) 1. 1 Therefore, we have ( ) a1 + a + a a > a 1 ( a + a a 1 ) 1 > ( a + a a ) 1 > 1 a ( a 3 + a a ) > a 3 ( a 4 + a a ) 3 > 3... a 1 + a > a 1 a Hece, by multiplicatio we get ( ) a1 + a +...a > (a 1 a...a ), or ( ) a1 + a +...a > (a 1 a...a ) 1. 50

57 We have give two proofs for this iequality for several reasos. I the first place, the iequality is iterestig i that it ca established i a large umber of ways. There are literally dozes of differet proofs for the arithmetic-geometric iequality that are based o ideas represetig a great variety of sources. I the secod place, it has a fudametal role i the theory of iequalities ad is the keystoe o which may other very importat results rest. I the third place, we ca use some of its cosequeces to solve a umber of maximizatio ad miimizatio problems (Beckebach 1961). Therefore, from the historical prospective ad its ubiquitousess i mathematics, this iequality truly deserve to be listed first. Just about every aspect of this iequality has bee ivestigated. Tug (1975) proved that the upper ad lower bouds of the differece betwee the arithmetic ad the geometric meas of quatities ca be give as follows: Theorem. Let x 1,x,...,x be positive real umbers. Suppose t ad T are the smallest ad largest values respectively of the give quatities, ad c = l[ T (T t) ]l T t l T t. The ( 1 ( T t) ) A G (ct + (1 c)t t c T 1 c ) We refer the reader to Tug (1975) for the proof of this theorem. Others mathematicias sought more exotic iequalities betwee the betwee the arithmetic ad geometric meas. For example, Kedlaya (1994) proved a cojecture made F. Hollad i 199 which states the followig: Theorem 3. Let x 1,x,...,x be positive real umbers. The arithmetic mea of the umbers x 1, x 1 x, 3 x 1 x x 3,..., x 1 x...x 51

58 does ot exceed the geometric mea of the umbers x 1, x 1x, x 1x x 3,..., x 1x...x 3 Now we establish the iequality betwee G ad H. Theorem 4. Let x 1,x,...,x be positive real umbers. The H G, ad H = G if ad oly if all the x i s are equal. Proof. Pick a 1,a,...,a to be positive real umbers. For the umbers a 1,a,...,a, by the arithmetic-geometric meas iequality, we have ( ) 1 a1 + a a 1 a 1 a a The 1 a 1 + a a 1 (a 1 a,...,a ) 1. Equality holds here if ad oly if a 1 1 = a 1 =... = a 1. Note by theorems 1 ad 4, we have: H G A, ad H = G = A if ad oly if all the x i s are equal. The followig theorem establishes a geeralized iequality betwee power meas. Theorem 5. Suppose p ad q are ozero real umbers such that p > q. The for ay positive umbers a 1,a,...,a, ( ap 1 + ap ap ) 1 p ( aq 1 + aq aq ) 1 q, 5

59 with equality holdig if ad oly if a 1 = a... = a. We preset the proof give by Schaumberger (1988). Proof. Suppose p > q > 0 or 0 > p > q, ad let f (x) = qx p + (p q) px q. The for x > 0: qx p + p q px q (4.18) with equality holdig if ad oly if x = 1. This follows from f (x) = qx p + (p q) px q has a absolute miimum at x = 1 (because f (x) = qpx p 1 qpx q 1 vaishes if ad oly if x = 1, ad f (1) = (qp(p q) > 0). Now cosider ay positive umbers a 1,a,...,a ad let A = ( aq 1 + aq aq ) 1 q. (4.19) Substitutig x i = a i A for 1 i i equatio 4.19 ad addig we get q( ap 1 +ap +...+ap A p ) + p q p( aq 1 +aq +...+aq A q ). Sice a q 1 + aq aq = A q, it follows that q( ap 1 + ap ap A p ) q (4.0) If p > q > 0, the equatio 4.19 gives a p 1 + ap ap A p, Which ca be writte as ( ap 1 + ap ap ) 1 p ( aq 1 + aq aq ) 1 q. If 0 > p > q, the divisio by q reverses the iequality ad equatio 4.19reduces to 53

60 a p 1 + ap ap A p. Raisig both sides to the 1 p power reverses the iequality agai, ad we have ( ap 1 + ap ap ) 1 p ( aq 1 + aq aq ) 1 q. Note, furthermore, that equality holds if ad oly if each of the substituted values x i = a i A which is equivalet to a 1 = a... = a. equals 1, Letp > 0 > q. The for x > 0, f (x) = qx p +(p q) px q has a absolute maximum at x = 1. Thus qx p + (p q) px q (4.1) with equality holdig if ad oly if x = 1. Agai let A = ( aq 1 + aq aq ) 1 q. Substitutig x i = a i A successively i i equatio 4.1, ad addig the iequalities, we obtai q( ap 1 + ap ap A p ) + p q p( aq 1 + aq aq A q ) = p. Thus, q( ap 1 + ap ap A p ) q. Sice 0 > q, divisio by q reverses the iequality ad leads to ( ap 1 + ap ap ) 1 p ( aq 1 + aq aq ) 1 q. 54

61 Equality holds if ad oly if each of the substituted values x i = a i A a 1 = a... = a equals 1, which is equivalet to 55

62 Chapter 5. Coclusio The history of meas is log ad lade with details as meas are used by ordiary people, experts, academicias, ad scietists to express a represetative umber that typifies a set of values. The atique meas appear to have bee well kow by the daw of Greek mathematics. After the 4th cetury A.D., the theory of meas appears to have come to a stadstill. With Pappus s documetatio of the te kow meas of his time i the 4th cetury A.D., it seems that the iterest i the theory of meas had waed. Boethius is the last kow aciet writer to metio the eleve classical meas as part of his work "Arithmetic", which is a commetary o works of Nicomachus ad Pappus (Gow 193). The ext referece to the classical meas after Boethius appeared as a quotatio by Ocreatus i his work "Prolgus I Helceph," which was writte i the 1 th or 13 th cetury A.D. (Heath 191). Thus, it seems reasoable to speculate that the oly meas that may have draw ay iterest to be metioed i mathematics literature betwee the 5 th cetury ad the 16 th cetury are the three atique meas which were treated as the oly types of meas. That the atique meas cotiued to be of iterest from the 5 th cetury to the 16 th cetury comes as o surprise, sice the arithmetic mea is ofte useful i commercial trasactios, the geometric mea was preserved by the use of the mea proportioal i geometry, ad the harmoic mea is closely tied to music theory. I Middle Ages, the golde sectio became a favorite topic of theological speculatio. May leared people, ispired by the argumets of the Pythagorias ad Platoists, sought ad foud i this proportio a key to the mystery of creatio, declarig that extreme ad mea ratios were the very priciple which the Supreme Architect had adopted i the cosmic ad global desig; hece, the title divie proportio bestowed upo this ratio (Datzig 1955). The late reaissace iterest i sciece ad aciet Greek mathematics brought about a ew iterest i the theory of meas as particular aspects of some physical pheomea ca best be expressed by usig their mea values. Advaces i the theory of statistics has show that the 56

63 arithmetic mea ca be used as a represetative of a set of observatio data (Buhler 1944). This outlook carried over quite aturally to fidig ew ways to express a mea of a set of values that best fits the purpose at had, ad thus theory of meas was rebor. This rich history motivates us to dig deeper ito the cocept of meas to uderstad the uderlyig foudatio i which it is couched. We close by otig that there is a proliferatio of meas i mathematics of which we oly touched o a few. This proliferatio may be attributed to the fact that ew ways to express a mea of a set of umbers arise i applicatios cotiuously. Ufortuately, our discussio did ot iclude may other types of meas that play a pivotal role i mathematics research such as the rich area of iterated meas, e.g. Gauss arithmetic-geometric mea, o-symmetric meas, ad weighted meas. Meas are coected with diverse areas of mathematics research from Fourier Series, error measuremets, to aggregatio ad social choice. With each ew mea developed, ways have to be foud that will relate this mea to the oes already kow. 57

64 Refereces [1] Allema, G.j.,Greek Geometry from Thales to Euclid, Hodges ad Figgs, [] Almkvist,G.ad Berdt,B. Gauss, Lade, the Arithmetic Mea, Ellipses, π, ad the Ladies Diaries.Ameri. math. Mothly, 95(1988),p [3] Beckebach E.F.A Class of Mea Value Fuctios. Ameri. math. Mothly, 57 (1950),p [4] Beckebach,E.F. ad Bellma, R., A itroductio to Iequalities, New York, Radom House, [5] Beetle, R. D. O the complete idepedece of Schimmack s postulate for the arithmetic mea, Mathematische Aale, Vol.76, 1915, p [6] Beetle, R. D. Bulleti of the America Mathematical Society, Vol., 1916, p [7] Bellma, R., Iequalities. Mathematics Magazie, Vol.8, 1954, p.1-6. [8] Bellma, R., O the arithmetic-geometric mea iequality. Mathematics Studet,Vol.4,1956/57, p [9] Bema, W. W. ad Smith, D.G., A Brief History of Mathematics, Chicago: Ope Court Publishig co., [10] Berlighoff, W. ad Gouvea, F., Math Through the Ages, Farmigto: Oxto House Publishers, 00. [11] Borwei, J. ad Borwei, P., Pi ad the AGM, New York: Wiley Itersciece, [1] Buhler W.K., Gauss: A Biographical Study. New York: Spriger, 1944,p [13] Bulle, P.S., Mitriovic, D.S., ad Vasic, P.M.,Meas ad Their Iequalities, Bosto: D. Reidel Publishig Compay,198. [14] Burrows, B.L. ad Talbot, R.F. Which Mea Do you Mea. It. J. Math. Educ. Sci. Techol., 1986, vol. 3, p [15] Carlso, B.C., Algorithms Ivolvig Arithmetic Ad Geometric Meas, Ameri. math. Mothly, 78(1971),p [16] Carlso, B.C., The Logarithmic Mea,Ameri. math. Mothly, 79(197),p [17] Chrystal, G., Algebra, Lodo: A ad C Black, 1916, p [18] Cox, D.A. The Arithmetic-Geometric Mea Of Gauss, L Eseigemet Mathematique, t.30 (1984), p [19] Datzig, T., Number The Laguage of Sciece Mathematics, New York: Macmilla,

65 [0] Datzig, T., The Bequest of the Greeks, Lodo: George Alle & Uwi LTD., [1] De Fietti, B., Sul Cocetto di media, Giorale dell istituto Italiao degli Attuari, Vol., 1931, p [] Dodd, E.L., The Covergece of Geeral Meas ad the Ivariace of Form of Certai Frequecy Fuctios. The America Joural of Mathematics, Vol. 49, No., (Apr., 197), p [3] Dodd, E.L., The complete Idepedece of Certai Properties of Meas. Aals of Mathematics, vol. 35,No.4,October, [4] Ercolao, J.L., Geometrical Iterpretatio of some classical iequalities. Math. Magazie. Vol. 45, 197, p [5] Ercolao, J.L., Remarks o the Neglected Mea. Mathematics Teacher. Vol. 66, 1973, p [6] Eves, H., Meas appearig i geometric figures. Math. Magazie, Vol. 76, No. 4, Oct. 003, p [7] Gallat, C., Proof Without Words:A truly geometric iequality. Mathematics Magazie, vol. 50,1977,p. 98. [8] Gow, J., A short History of Greek Mathematics, New York: G.E. Stechert & co., 193. [9] Gratta-Guiess, I, The Search for Mathematical Roots, , Piceto: Priceto Uiversity Press, 000. [30] Hald, A., A History of Mathematical Statistics from 1750 to 1930, New York: Joh Wiley & Sos, Publicatio [31] Heath,T., A History of Greek Mathematics, Vol. 1, Lodo: Oxford Uiversity Press, 191. [3] Heath, T. L., A Maual of Greek Mathematics, Dover Publicatio, New York, [33] Hoeh, L. ad Nive, I., Averages o the Move, Mathematics Magazie, Vol. 58, No. 3, (May, 1985),p [34] Hubbard, J. ad Moll, V., A geometric view of ratioal Lade trasformatio, Bull. Lodo Math. Soc., Vol. 35, 003, p [35] Hutigto, E.V., Sets of Idepedet Postulates for the Arithmetic Mea, the Geometric Mea, the Harmoic Mea, ad the the Root-Mea-Square. Trasactios of the America Mathematical Society, Vol. 9,No.1 (Ja. 197), p 1-. [36] Kedlaya, K., Proof of a mixed Arithmetic-mea, Geometric-Mea Iequality. America Mathematical Mothly, Vol. 101, No. 4 (Apr., 1994),p

66 [37] Klei, J. Greek Mathematical Thought ad the Origi of Algebra, Cambridge: M. I. T. Press, [38] Koold, C. ad Pollatsek, A., Data Aalysis as the Search for sigals i oisy processes. Joural for research i mathematics educatio, 00, vol. 33,p [39] Lagto, S., Gauss, recurrece relatio, ad the AGM, Talk delivered at sessio Mathematics i the Age of Euler. Joit Mathematics Meetigs, New Orleas, Louisiaa, 1/11/001. [40] Leach, E. b. ad Sholader, M.C., Exteded Mea Values, America Mathematical Mothly, Vol. 85, No. (Feb., 1978),p [41] Lewiter, M. ad Widulski, W., The Saga of Mathematics,Upper Saddle River: Pretice Hall, 00. [4] Li, Tug-Po The Power Mea ad the Logarithmic Mea, Ameri. math. Mothly, 81(1974),p [43] Madde, J., Some Notes o Meas, Upublished paper, 6/3/000. [44] Mays, M.E., Fuctios Which Parametrize Meas. The America Mathematical Mothly, Vol. 90, No.10, (Dec., 1983), p [45] Maziarz, E.A. ad Greewood, T., Greek Mathematical Philosophy, New York: Frederick Ugar Publishig co., [46] Moroey, M.J., O the Average ad Scatter. The World of Mathematics. Newma, J.R., Editor, Simo ad Schuster,New York, 1956, p [47] Moskovitz, D., A Aligmet Chart for Various Meas. The America Mathematical Mothly, Vol. 40, No.10, (Dec., 1933), p [48] Mosteller, F. ad Rourke, R., Probability with Statistical Applicatios, Lodo:Addiso-Wesley Publishig Co [49] Nagumo, M., Uber eie Klasse der Mittelwerte, Japaese Joural of Mathematics, Vol. 7, 1930, p [50] Oxford Eglish Dictioary, Oxford Uiversity Press, Oxford: 004. [51] Pitteger, A.O., The Logarithmic Mea i Variables, America Mathematical Mothly, Vol. 9, No. (Feb., 1985),p [5] Polya, G., O the Harmoic mea of two umbers, America Math. Moth., vol.57, 1960, p.6-8. Polya, G., The miimum fractio of the popular vote that ca elect the presidet of the Uited States. Mathematics Teacher, vol.54, 1961, p

67 [53] Ruthig, D., Proofs of the arithmetic-geometric mea iequality, It. J. Math. Educ. Sci. Techol., 198, vol. 13, p [54] Shaughsey, J.,ad Bergma, B., Thikig about Ucertaity, Wilso, P. editor,research ideas for the classroom, Macmilla New York, 1993,p [55] Schaumberger, N., Aother Proof of the Iequality betwee Power Meas. The College Mathematics Joural, Vol. 19, No. 1 (Ja., 1988) p [56] Schild, A., Geometry of the Meas. Mathematics Teacher. Vol. 67, 1974, p [57] Schimmack, R., Der Satz vom arithmetische Mittel i axiomatischer Begrudug. Mathematische Aale, vol. 68, 1909, p [58] Schoeberg, I. J., O the Arithmetic-Geometric Mea, Delta,vol.7, Fall 1977, p [59] Schoeberg, I. J., O the Arithmetic-Geometric Mea ad similar iterative algorithms, Mathematical Time Exposures, Mathematical Associatio of America, 198, p [60] Scott,W.R., Meas i Groups,America Joural of Mathematics, Vol. 74, No.3, (Jul., 195), p [61] Slay, J.C. ad Solomo, J.L.,A Mea Geeratig Fuctio, The Two-Year College Mathematics Joural, Vol. 1, No.1, (Ja., 1981),7-9. [6] Smith, D.E., History of Mathematics, Vol., New York: Dover,1951. [63] Steves, S.S., O the Averagig of Data, Sciece, Vol.11, No.3135, (Ja. 8, 1955), p [64] Sulliva, J.J., The electio of a presidet. Mathematics Teacher, vol.65, 197, p [65] Tag, J. O the costructio ad iterpretatio of meas, It. J. Math. Educ. Sci. Techol., 1983, Vol. 14, No.1, p [66] Thacker, A., Demostratio of the kow Theorem, that the Arithmetic Mea betwee ay umber of positive quatities is greater tha their geometric. Cambridge ad Dubli mathematical Joural, vol. 6, 1851, p.81=-83. [67] Thomas, I., Selectios Illustratig the History of Greek Mathematics, Vol.1;,Cambridge, Harvard Press, [68] Tug, S. H., O the lower ad upper bouds of the differece betwee the Arithmetic-Geometric Mea. Math. of Computatio,vol.9,July 1975, p [69] Vamaamurthy, M.K., Iequalities for Meas. Joural of Mathematical Aalysis ad Applicatios, Vol. 183, (1994), p [70] Watso, G.N., The Marquis ad the lad-aget. Math. Gaz., vol.17,1933, p

68 [71] Yu, Li-We, Some Geeralizatios of the Iequalities about Mea, Tamsui Oxford Jour. of Math. Sc.,Vol.18, 00, p

69 Vita Mabrouck Khalifa Faradj was bor o February 5, 1958, i Fort Lamy (ow N Djamea), Chad. He fiished his udergraduate studies at Louisiaa State Uiversity i May, He eared the degree of Master of Public Admiistratio from Louisiaa State Uiversity i August, He is curretly a cadidate for the degree of Master of Sciece i mathematics, which will be awarded i August

Trigonometric Form of a Complex Number. The Complex Plane. axis. ( 2, 1) or 2 i FIGURE 6.44. The absolute value of the complex number z a bi is

Trigonometric Form of a Complex Number. The Complex Plane. axis. ( 2, 1) or 2 i FIGURE 6.44. The absolute value of the complex number z a bi is 0_0605.qxd /5/05 0:45 AM Page 470 470 Chapter 6 Additioal Topics i Trigoometry 6.5 Trigoometric Form of a Complex Number What you should lear Plot complex umbers i the complex plae ad fid absolute values

More information

Convexity, Inequalities, and Norms

Convexity, Inequalities, and Norms Covexity, Iequalities, ad Norms Covex Fuctios You are probably familiar with the otio of cocavity of fuctios. Give a twicedifferetiable fuctio ϕ: R R, We say that ϕ is covex (or cocave up) if ϕ (x) 0 for

More information

Sequences and Series

Sequences and Series CHAPTER 9 Sequeces ad Series 9.. Covergece: Defiitio ad Examples Sequeces The purpose of this chapter is to itroduce a particular way of geeratig algorithms for fidig the values of fuctios defied by their

More information

In nite Sequences. Dr. Philippe B. Laval Kennesaw State University. October 9, 2008

In nite Sequences. Dr. Philippe B. Laval Kennesaw State University. October 9, 2008 I ite Sequeces Dr. Philippe B. Laval Keesaw State Uiversity October 9, 2008 Abstract This had out is a itroductio to i ite sequeces. mai de itios ad presets some elemetary results. It gives the I ite Sequeces

More information

Department of Computer Science, University of Otago

Department of Computer Science, University of Otago Departmet of Computer Sciece, Uiversity of Otago Techical Report OUCS-2006-09 Permutatios Cotaiig May Patters Authors: M.H. Albert Departmet of Computer Sciece, Uiversity of Otago Micah Colema, Rya Fly

More information

I. Chi-squared Distributions

I. Chi-squared Distributions 1 M 358K Supplemet to Chapter 23: CHI-SQUARED DISTRIBUTIONS, T-DISTRIBUTIONS, AND DEGREES OF FREEDOM To uderstad t-distributios, we first eed to look at aother family of distributios, the chi-squared distributios.

More information

1. MATHEMATICAL INDUCTION

1. MATHEMATICAL INDUCTION 1. MATHEMATICAL INDUCTION EXAMPLE 1: Prove that for ay iteger 1. Proof: 1 + 2 + 3 +... + ( + 1 2 (1.1 STEP 1: For 1 (1.1 is true, sice 1 1(1 + 1. 2 STEP 2: Suppose (1.1 is true for some k 1, that is 1

More information

Properties of MLE: consistency, asymptotic normality. Fisher information.

Properties of MLE: consistency, asymptotic normality. Fisher information. Lecture 3 Properties of MLE: cosistecy, asymptotic ormality. Fisher iformatio. I this sectio we will try to uderstad why MLEs are good. Let us recall two facts from probability that we be used ofte throughout

More information

Factors of sums of powers of binomial coefficients

Factors of sums of powers of binomial coefficients ACTA ARITHMETICA LXXXVI.1 (1998) Factors of sums of powers of biomial coefficiets by Neil J. Cali (Clemso, S.C.) Dedicated to the memory of Paul Erdős 1. Itroductio. It is well ow that if ( ) a f,a = the

More information

Chapter 6: Variance, the law of large numbers and the Monte-Carlo method

Chapter 6: Variance, the law of large numbers and the Monte-Carlo method Chapter 6: Variace, the law of large umbers ad the Mote-Carlo method Expected value, variace, ad Chebyshev iequality. If X is a radom variable recall that the expected value of X, E[X] is the average value

More information

SAMPLE QUESTIONS FOR FINAL EXAM. (1) (2) (3) (4) Find the following using the definition of the Riemann integral: (2x + 1)dx

SAMPLE QUESTIONS FOR FINAL EXAM. (1) (2) (3) (4) Find the following using the definition of the Riemann integral: (2x + 1)dx SAMPLE QUESTIONS FOR FINAL EXAM REAL ANALYSIS I FALL 006 3 4 Fid the followig usig the defiitio of the Riema itegral: a 0 x + dx 3 Cosider the partitio P x 0 3, x 3 +, x 3 +,......, x 3 3 + 3 of the iterval

More information

.04. This means $1000 is multiplied by 1.02 five times, once for each of the remaining sixmonth

.04. This means $1000 is multiplied by 1.02 five times, once for each of the remaining sixmonth Questio 1: What is a ordiary auity? Let s look at a ordiary auity that is certai ad simple. By this, we mea a auity over a fixed term whose paymet period matches the iterest coversio period. Additioally,

More information

Chapter 7 Methods of Finding Estimators

Chapter 7 Methods of Finding Estimators Chapter 7 for BST 695: Special Topics i Statistical Theory. Kui Zhag, 011 Chapter 7 Methods of Fidig Estimators Sectio 7.1 Itroductio Defiitio 7.1.1 A poit estimator is ay fuctio W( X) W( X1, X,, X ) of

More information

FOUNDATIONS OF MATHEMATICS AND PRE-CALCULUS GRADE 10

FOUNDATIONS OF MATHEMATICS AND PRE-CALCULUS GRADE 10 FOUNDATIONS OF MATHEMATICS AND PRE-CALCULUS GRADE 10 [C] Commuicatio Measuremet A1. Solve problems that ivolve liear measuremet, usig: SI ad imperial uits of measure estimatio strategies measuremet strategies.

More information

A probabilistic proof of a binomial identity

A probabilistic proof of a binomial identity A probabilistic proof of a biomial idetity Joatho Peterso Abstract We give a elemetary probabilistic proof of a biomial idetity. The proof is obtaied by computig the probability of a certai evet i two

More information

Chapter 5: Inner Product Spaces

Chapter 5: Inner Product Spaces Chapter 5: Ier Product Spaces Chapter 5: Ier Product Spaces SECION A Itroductio to Ier Product Spaces By the ed of this sectio you will be able to uderstad what is meat by a ier product space give examples

More information

Soving Recurrence Relations

Soving Recurrence Relations Sovig Recurrece Relatios Part 1. Homogeeous liear 2d degree relatios with costat coefficiets. Cosider the recurrece relatio ( ) T () + at ( 1) + bt ( 2) = 0 This is called a homogeeous liear 2d degree

More information

THE ARITHMETIC OF INTEGERS. - multiplication, exponentiation, division, addition, and subtraction

THE ARITHMETIC OF INTEGERS. - multiplication, exponentiation, division, addition, and subtraction THE ARITHMETIC OF INTEGERS - multiplicatio, expoetiatio, divisio, additio, ad subtractio What to do ad what ot to do. THE INTEGERS Recall that a iteger is oe of the whole umbers, which may be either positive,

More information

MARTINGALES AND A BASIC APPLICATION

MARTINGALES AND A BASIC APPLICATION MARTINGALES AND A BASIC APPLICATION TURNER SMITH Abstract. This paper will develop the measure-theoretic approach to probability i order to preset the defiitio of martigales. From there we will apply this

More information

Asymptotic Growth of Functions

Asymptotic Growth of Functions CMPS Itroductio to Aalysis of Algorithms Fall 3 Asymptotic Growth of Fuctios We itroduce several types of asymptotic otatio which are used to compare the performace ad efficiecy of algorithms As we ll

More information

Case Study. Normal and t Distributions. Density Plot. Normal Distributions

Case Study. Normal and t Distributions. Density Plot. Normal Distributions Case Study Normal ad t Distributios Bret Halo ad Bret Larget Departmet of Statistics Uiversity of Wiscosi Madiso October 11 13, 2011 Case Study Body temperature varies withi idividuals over time (it ca

More information

where: T = number of years of cash flow in investment's life n = the year in which the cash flow X n i = IRR = the internal rate of return

where: T = number of years of cash flow in investment's life n = the year in which the cash flow X n i = IRR = the internal rate of return EVALUATING ALTERNATIVE CAPITAL INVESTMENT PROGRAMS By Ke D. Duft, Extesio Ecoomist I the March 98 issue of this publicatio we reviewed the procedure by which a capital ivestmet project was assessed. The

More information

CS103A Handout 23 Winter 2002 February 22, 2002 Solving Recurrence Relations

CS103A Handout 23 Winter 2002 February 22, 2002 Solving Recurrence Relations CS3A Hadout 3 Witer 00 February, 00 Solvig Recurrece Relatios Itroductio A wide variety of recurrece problems occur i models. Some of these recurrece relatios ca be solved usig iteratio or some other ad

More information

CHAPTER 3 THE TIME VALUE OF MONEY

CHAPTER 3 THE TIME VALUE OF MONEY CHAPTER 3 THE TIME VALUE OF MONEY OVERVIEW A dollar i the had today is worth more tha a dollar to be received i the future because, if you had it ow, you could ivest that dollar ad ear iterest. Of all

More information

Class Meeting # 16: The Fourier Transform on R n

Class Meeting # 16: The Fourier Transform on R n MATH 18.152 COUSE NOTES - CLASS MEETING # 16 18.152 Itroductio to PDEs, Fall 2011 Professor: Jared Speck Class Meetig # 16: The Fourier Trasform o 1. Itroductio to the Fourier Trasform Earlier i the course,

More information

Lecture 4: Cauchy sequences, Bolzano-Weierstrass, and the Squeeze theorem

Lecture 4: Cauchy sequences, Bolzano-Weierstrass, and the Squeeze theorem Lecture 4: Cauchy sequeces, Bolzao-Weierstrass, ad the Squeeze theorem The purpose of this lecture is more modest tha the previous oes. It is to state certai coditios uder which we are guarateed that limits

More information

The analysis of the Cournot oligopoly model considering the subjective motive in the strategy selection

The analysis of the Cournot oligopoly model considering the subjective motive in the strategy selection The aalysis of the Courot oligopoly model cosiderig the subjective motive i the strategy selectio Shigehito Furuyama Teruhisa Nakai Departmet of Systems Maagemet Egieerig Faculty of Egieerig Kasai Uiversity

More information

Building Blocks Problem Related to Harmonic Series

Building Blocks Problem Related to Harmonic Series TMME, vol3, o, p.76 Buildig Blocks Problem Related to Harmoic Series Yutaka Nishiyama Osaka Uiversity of Ecoomics, Japa Abstract: I this discussio I give a eplaatio of the divergece ad covergece of ifiite

More information

SEQUENCES AND SERIES

SEQUENCES AND SERIES Chapter 9 SEQUENCES AND SERIES Natural umbers are the product of huma spirit. DEDEKIND 9.1 Itroductio I mathematics, the word, sequece is used i much the same way as it is i ordiary Eglish. Whe we say

More information

1 Computing the Standard Deviation of Sample Means

1 Computing the Standard Deviation of Sample Means Computig the Stadard Deviatio of Sample Meas Quality cotrol charts are based o sample meas ot o idividual values withi a sample. A sample is a group of items, which are cosidered all together for our aalysis.

More information

Discrete Mathematics and Probability Theory Spring 2014 Anant Sahai Note 13

Discrete Mathematics and Probability Theory Spring 2014 Anant Sahai Note 13 EECS 70 Discrete Mathematics ad Probability Theory Sprig 2014 Aat Sahai Note 13 Itroductio At this poit, we have see eough examples that it is worth just takig stock of our model of probability ad may

More information

Week 3 Conditional probabilities, Bayes formula, WEEK 3 page 1 Expected value of a random variable

Week 3 Conditional probabilities, Bayes formula, WEEK 3 page 1 Expected value of a random variable Week 3 Coditioal probabilities, Bayes formula, WEEK 3 page 1 Expected value of a radom variable We recall our discussio of 5 card poker hads. Example 13 : a) What is the probability of evet A that a 5

More information

3. Greatest Common Divisor - Least Common Multiple

3. Greatest Common Divisor - Least Common Multiple 3 Greatest Commo Divisor - Least Commo Multiple Defiitio 31: The greatest commo divisor of two atural umbers a ad b is the largest atural umber c which divides both a ad b We deote the greatest commo gcd

More information

Notes on exponential generating functions and structures.

Notes on exponential generating functions and structures. Notes o expoetial geeratig fuctios ad structures. 1. The cocept of a structure. Cosider the followig coutig problems: (1) to fid for each the umber of partitios of a -elemet set, (2) to fid for each the

More information

FIBONACCI NUMBERS: AN APPLICATION OF LINEAR ALGEBRA. 1. Powers of a matrix

FIBONACCI NUMBERS: AN APPLICATION OF LINEAR ALGEBRA. 1. Powers of a matrix FIBONACCI NUMBERS: AN APPLICATION OF LINEAR ALGEBRA. Powers of a matrix We begi with a propositio which illustrates the usefuless of the diagoalizatio. Recall that a square matrix A is diogaalizable if

More information

Modified Line Search Method for Global Optimization

Modified Line Search Method for Global Optimization Modified Lie Search Method for Global Optimizatio Cria Grosa ad Ajith Abraham Ceter of Excellece for Quatifiable Quality of Service Norwegia Uiversity of Sciece ad Techology Trodheim, Norway {cria, ajith}@q2s.tu.o

More information

SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES

SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES SECTION 1.5 : SUMMATION NOTATION + WORK WITH SEQUENCES Read Sectio 1.5 (pages 5 9) Overview I Sectio 1.5 we lear to work with summatio otatio ad formulas. We will also itroduce a brief overview of sequeces,

More information

Basic Elements of Arithmetic Sequences and Series

Basic Elements of Arithmetic Sequences and Series MA40S PRE-CALCULUS UNIT G GEOMETRIC SEQUENCES CLASS NOTES (COMPLETED NO NEED TO COPY NOTES FROM OVERHEAD) Basic Elemets of Arithmetic Sequeces ad Series Objective: To establish basic elemets of arithmetic

More information

1 Correlation and Regression Analysis

1 Correlation and Regression Analysis 1 Correlatio ad Regressio Aalysis I this sectio we will be ivestigatig the relatioship betwee two cotiuous variable, such as height ad weight, the cocetratio of a ijected drug ad heart rate, or the cosumptio

More information

hp calculators HP 12C Statistics - average and standard deviation Average and standard deviation concepts HP12C average and standard deviation

hp calculators HP 12C Statistics - average and standard deviation Average and standard deviation concepts HP12C average and standard deviation HP 1C Statistics - average ad stadard deviatio Average ad stadard deviatio cocepts HP1C average ad stadard deviatio Practice calculatig averages ad stadard deviatios with oe or two variables HP 1C Statistics

More information

Here are a couple of warnings to my students who may be here to get a copy of what happened on a day that you missed.

Here are a couple of warnings to my students who may be here to get a copy of what happened on a day that you missed. This documet was writte ad copyrighted by Paul Dawkis. Use of this documet ad its olie versio is govered by the Terms ad Coditios of Use located at http://tutorial.math.lamar.edu/terms.asp. The olie versio

More information

THE HEIGHT OF q-binary SEARCH TREES

THE HEIGHT OF q-binary SEARCH TREES THE HEIGHT OF q-binary SEARCH TREES MICHAEL DRMOTA AND HELMUT PRODINGER Abstract. q biary search trees are obtaied from words, equipped with the geometric distributio istead of permutatios. The average

More information

Section 11.3: The Integral Test

Section 11.3: The Integral Test Sectio.3: The Itegral Test Most of the series we have looked at have either diverged or have coverged ad we have bee able to fid what they coverge to. I geeral however, the problem is much more difficult

More information

Taking DCOP to the Real World: Efficient Complete Solutions for Distributed Multi-Event Scheduling

Taking DCOP to the Real World: Efficient Complete Solutions for Distributed Multi-Event Scheduling Taig DCOP to the Real World: Efficiet Complete Solutios for Distributed Multi-Evet Schedulig Rajiv T. Maheswara, Milid Tambe, Emma Bowrig, Joatha P. Pearce, ad Pradeep araatham Uiversity of Souther Califoria

More information

Our aim is to show that under reasonable assumptions a given 2π-periodic function f can be represented as convergent series

Our aim is to show that under reasonable assumptions a given 2π-periodic function f can be represented as convergent series 8 Fourier Series Our aim is to show that uder reasoable assumptios a give -periodic fuctio f ca be represeted as coverget series f(x) = a + (a cos x + b si x). (8.) By defiitio, the covergece of the series

More information

How To Solve The Homewor Problem Beautifully

How To Solve The Homewor Problem Beautifully Egieerig 33 eautiful Homewor et 3 of 7 Kuszmar roblem.5.5 large departmet store sells sport shirts i three sizes small, medium, ad large, three patters plaid, prit, ad stripe, ad two sleeve legths log

More information

AP Calculus BC 2003 Scoring Guidelines Form B

AP Calculus BC 2003 Scoring Guidelines Form B AP Calculus BC Scorig Guidelies Form B The materials icluded i these files are iteded for use by AP teachers for course ad exam preparatio; permissio for ay other use must be sought from the Advaced Placemet

More information

Theorems About Power Series

Theorems About Power Series Physics 6A Witer 20 Theorems About Power Series Cosider a power series, f(x) = a x, () where the a are real coefficiets ad x is a real variable. There exists a real o-egative umber R, called the radius

More information

INVESTMENT PERFORMANCE COUNCIL (IPC)

INVESTMENT PERFORMANCE COUNCIL (IPC) INVESTMENT PEFOMANCE COUNCIL (IPC) INVITATION TO COMMENT: Global Ivestmet Performace Stadards (GIPS ) Guidace Statemet o Calculatio Methodology The Associatio for Ivestmet Maagemet ad esearch (AIM) seeks

More information

Example 2 Find the square root of 0. The only square root of 0 is 0 (since 0 is not positive or negative, so those choices don t exist here).

Example 2 Find the square root of 0. The only square root of 0 is 0 (since 0 is not positive or negative, so those choices don t exist here). BEGINNING ALGEBRA Roots ad Radicals (revised summer, 00 Olso) Packet to Supplemet the Curret Textbook - Part Review of Square Roots & Irratioals (This portio ca be ay time before Part ad should mostly

More information

2-3 The Remainder and Factor Theorems

2-3 The Remainder and Factor Theorems - The Remaider ad Factor Theorems Factor each polyomial completely usig the give factor ad log divisio 1 x + x x 60; x + So, x + x x 60 = (x + )(x x 15) Factorig the quadratic expressio yields x + x x

More information

INFINITE SERIES KEITH CONRAD

INFINITE SERIES KEITH CONRAD INFINITE SERIES KEITH CONRAD. Itroductio The two basic cocepts of calculus, differetiatio ad itegratio, are defied i terms of limits (Newto quotiets ad Riema sums). I additio to these is a third fudametal

More information

PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM

PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM PROCEEDINGS OF THE YEREVAN STATE UNIVERSITY Physical ad Mathematical Scieces 2015, 1, p. 15 19 M a t h e m a t i c s AN ALTERNATIVE MODEL FOR BONUS-MALUS SYSTEM A. G. GULYAN Chair of Actuarial Mathematics

More information

CHAPTER 7: Central Limit Theorem: CLT for Averages (Means)

CHAPTER 7: Central Limit Theorem: CLT for Averages (Means) CHAPTER 7: Cetral Limit Theorem: CLT for Averages (Meas) X = the umber obtaied whe rollig oe six sided die oce. If we roll a six sided die oce, the mea of the probability distributio is X P(X = x) Simulatio:

More information

Biology 171L Environment and Ecology Lab Lab 2: Descriptive Statistics, Presenting Data and Graphing Relationships

Biology 171L Environment and Ecology Lab Lab 2: Descriptive Statistics, Presenting Data and Graphing Relationships Biology 171L Eviromet ad Ecology Lab Lab : Descriptive Statistics, Presetig Data ad Graphig Relatioships Itroductio Log lists of data are ofte ot very useful for idetifyig geeral treds i the data or the

More information

5: Introduction to Estimation

5: Introduction to Estimation 5: Itroductio to Estimatio Cotets Acroyms ad symbols... 1 Statistical iferece... Estimatig µ with cofidece... 3 Samplig distributio of the mea... 3 Cofidece Iterval for μ whe σ is kow before had... 4 Sample

More information

How Euler Did It. In a more modern treatment, Hardy and Wright [H+W] state this same theorem as. n n+ is perfect.

How Euler Did It. In a more modern treatment, Hardy and Wright [H+W] state this same theorem as. n n+ is perfect. Amicable umbers November 005 How Euler Did It by Ed Sadifer Six is a special umber. It is divisible by, ad 3, ad, i what at first looks like a strage coicidece, 6 = + + 3. The umber 8 shares this remarkable

More information

THE ABRACADABRA PROBLEM

THE ABRACADABRA PROBLEM THE ABRACADABRA PROBLEM FRANCESCO CARAVENNA Abstract. We preset a detailed solutio of Exercise E0.6 i [Wil9]: i a radom sequece of letters, draw idepedetly ad uiformly from the Eglish alphabet, the expected

More information

NATIONAL SENIOR CERTIFICATE GRADE 12

NATIONAL SENIOR CERTIFICATE GRADE 12 NATIONAL SENIOR CERTIFICATE GRADE MATHEMATICS P EXEMPLAR 04 MARKS: 50 TIME: 3 hours This questio paper cosists of 8 pages ad iformatio sheet. Please tur over Mathematics/P DBE/04 NSC Grade Eemplar INSTRUCTIONS

More information

THE REGRESSION MODEL IN MATRIX FORM. For simple linear regression, meaning one predictor, the model is. for i = 1, 2, 3,, n

THE REGRESSION MODEL IN MATRIX FORM. For simple linear regression, meaning one predictor, the model is. for i = 1, 2, 3,, n We will cosider the liear regressio model i matrix form. For simple liear regressio, meaig oe predictor, the model is i = + x i + ε i for i =,,,, This model icludes the assumptio that the ε i s are a sample

More information

Maximum Likelihood Estimators.

Maximum Likelihood Estimators. Lecture 2 Maximum Likelihood Estimators. Matlab example. As a motivatio, let us look at oe Matlab example. Let us geerate a radom sample of size 00 from beta distributio Beta(5, 2). We will lear the defiitio

More information

1. C. The formula for the confidence interval for a population mean is: x t, which was

1. C. The formula for the confidence interval for a population mean is: x t, which was s 1. C. The formula for the cofidece iterval for a populatio mea is: x t, which was based o the sample Mea. So, x is guarateed to be i the iterval you form.. D. Use the rule : p-value

More information

Definition. A variable X that takes on values X 1, X 2, X 3,...X k with respective frequencies f 1, f 2, f 3,...f k has mean

Definition. A variable X that takes on values X 1, X 2, X 3,...X k with respective frequencies f 1, f 2, f 3,...f k has mean 1 Social Studies 201 October 13, 2004 Note: The examples i these otes may be differet tha used i class. However, the examples are similar ad the methods used are idetical to what was preseted i class.

More information

Chapter 7 - Sampling Distributions. 1 Introduction. What is statistics? It consist of three major areas:

Chapter 7 - Sampling Distributions. 1 Introduction. What is statistics? It consist of three major areas: Chapter 7 - Samplig Distributios 1 Itroductio What is statistics? It cosist of three major areas: Data Collectio: samplig plas ad experimetal desigs Descriptive Statistics: umerical ad graphical summaries

More information

5.4 Amortization. Question 1: How do you find the present value of an annuity? Question 2: How is a loan amortized?

5.4 Amortization. Question 1: How do you find the present value of an annuity? Question 2: How is a loan amortized? 5.4 Amortizatio Questio 1: How do you fid the preset value of a auity? Questio 2: How is a loa amortized? Questio 3: How do you make a amortizatio table? Oe of the most commo fiacial istrumets a perso

More information

BINOMIAL EXPANSIONS 12.5. In this section. Some Examples. Obtaining the Coefficients

BINOMIAL EXPANSIONS 12.5. In this section. Some Examples. Obtaining the Coefficients 652 (12-26) Chapter 12 Sequeces ad Series 12.5 BINOMIAL EXPANSIONS I this sectio Some Examples Otaiig the Coefficiets The Biomial Theorem I Chapter 5 you leared how to square a iomial. I this sectio you

More information

Ekkehart Schlicht: Economic Surplus and Derived Demand

Ekkehart Schlicht: Economic Surplus and Derived Demand Ekkehart Schlicht: Ecoomic Surplus ad Derived Demad Muich Discussio Paper No. 2006-17 Departmet of Ecoomics Uiversity of Muich Volkswirtschaftliche Fakultät Ludwig-Maximilias-Uiversität Müche Olie at http://epub.ub.ui-mueche.de/940/

More information

Chapter 5 O A Cojecture Of Erdíos Proceedigs NCUR VIII è1994è, Vol II, pp 794í798 Jeærey F Gold Departmet of Mathematics, Departmet of Physics Uiversity of Utah Do H Tucker Departmet of Mathematics Uiversity

More information

Output Analysis (2, Chapters 10 &11 Law)

Output Analysis (2, Chapters 10 &11 Law) B. Maddah ENMG 6 Simulatio 05/0/07 Output Aalysis (, Chapters 10 &11 Law) Comparig alterative system cofiguratio Sice the output of a simulatio is radom, the comparig differet systems via simulatio should

More information

Elementary Theory of Russian Roulette

Elementary Theory of Russian Roulette Elemetary Theory of Russia Roulette -iterestig patters of fractios- Satoshi Hashiba Daisuke Miematsu Ryohei Miyadera Itroductio. Today we are goig to study mathematical theory of Russia roulette. If some

More information

3 Basic Definitions of Probability Theory

3 Basic Definitions of Probability Theory 3 Basic Defiitios of Probability Theory 3defprob.tex: Feb 10, 2003 Classical probability Frequecy probability axiomatic probability Historical developemet: Classical Frequecy Axiomatic The Axiomatic defiitio

More information

Incremental calculation of weighted mean and variance

Incremental calculation of weighted mean and variance Icremetal calculatio of weighted mea ad variace Toy Fich [email protected] [email protected] Uiversity of Cambridge Computig Service February 009 Abstract I these otes I eplai how to derive formulae for umerically

More information

WHEN IS THE (CO)SINE OF A RATIONAL ANGLE EQUAL TO A RATIONAL NUMBER?

WHEN IS THE (CO)SINE OF A RATIONAL ANGLE EQUAL TO A RATIONAL NUMBER? WHEN IS THE (CO)SINE OF A RATIONAL ANGLE EQUAL TO A RATIONAL NUMBER? JÖRG JAHNEL 1. My Motivatio Some Sort of a Itroductio Last term I tought Topological Groups at the Göttige Georg August Uiversity. This

More information

Approximating Area under a curve with rectangles. To find the area under a curve we approximate the area using rectangles and then use limits to find

Approximating Area under a curve with rectangles. To find the area under a curve we approximate the area using rectangles and then use limits to find 1.8 Approximatig Area uder a curve with rectagles 1.6 To fid the area uder a curve we approximate the area usig rectagles ad the use limits to fid 1.4 the area. Example 1 Suppose we wat to estimate 1.

More information

Confidence Intervals for One Mean

Confidence Intervals for One Mean Chapter 420 Cofidece Itervals for Oe Mea Itroductio This routie calculates the sample size ecessary to achieve a specified distace from the mea to the cofidece limit(s) at a stated cofidece level for a

More information

Your organization has a Class B IP address of 166.144.0.0 Before you implement subnetting, the Network ID and Host ID are divided as follows:

Your organization has a Class B IP address of 166.144.0.0 Before you implement subnetting, the Network ID and Host ID are divided as follows: Subettig Subettig is used to subdivide a sigle class of etwork i to multiple smaller etworks. Example: Your orgaizatio has a Class B IP address of 166.144.0.0 Before you implemet subettig, the Network

More information

Chapter 5 Unit 1. IET 350 Engineering Economics. Learning Objectives Chapter 5. Learning Objectives Unit 1. Annual Amount and Gradient Functions

Chapter 5 Unit 1. IET 350 Engineering Economics. Learning Objectives Chapter 5. Learning Objectives Unit 1. Annual Amount and Gradient Functions Chapter 5 Uit Aual Amout ad Gradiet Fuctios IET 350 Egieerig Ecoomics Learig Objectives Chapter 5 Upo completio of this chapter you should uderstad: Calculatig future values from aual amouts. Calculatig

More information

CS103X: Discrete Structures Homework 4 Solutions

CS103X: Discrete Structures Homework 4 Solutions CS103X: Discrete Structures Homewor 4 Solutios Due February 22, 2008 Exercise 1 10 poits. Silico Valley questios: a How may possible six-figure salaries i whole dollar amouts are there that cotai at least

More information

Chapter 7: Confidence Interval and Sample Size

Chapter 7: Confidence Interval and Sample Size Chapter 7: Cofidece Iterval ad Sample Size Learig Objectives Upo successful completio of Chapter 7, you will be able to: Fid the cofidece iterval for the mea, proportio, ad variace. Determie the miimum

More information

Entropy of bi-capacities

Entropy of bi-capacities Etropy of bi-capacities Iva Kojadiovic LINA CNRS FRE 2729 Site école polytechique de l uiv. de Nates Rue Christia Pauc 44306 Nates, Frace [email protected] Jea-Luc Marichal Applied Mathematics

More information

0.7 0.6 0.2 0 0 96 96.5 97 97.5 98 98.5 99 99.5 100 100.5 96.5 97 97.5 98 98.5 99 99.5 100 100.5

0.7 0.6 0.2 0 0 96 96.5 97 97.5 98 98.5 99 99.5 100 100.5 96.5 97 97.5 98 98.5 99 99.5 100 100.5 Sectio 13 Kolmogorov-Smirov test. Suppose that we have a i.i.d. sample X 1,..., X with some ukow distributio P ad we would like to test the hypothesis that P is equal to a particular distributio P 0, i.e.

More information

Permutations, the Parity Theorem, and Determinants

Permutations, the Parity Theorem, and Determinants 1 Permutatios, the Parity Theorem, ad Determiats Joh A. Guber Departmet of Electrical ad Computer Egieerig Uiversity of Wiscosi Madiso Cotets 1 What is a Permutatio 1 2 Cycles 2 2.1 Traspositios 4 3 Orbits

More information

Hypothesis testing. Null and alternative hypotheses

Hypothesis testing. Null and alternative hypotheses Hypothesis testig Aother importat use of samplig distributios is to test hypotheses about populatio parameters, e.g. mea, proportio, regressio coefficiets, etc. For example, it is possible to stipulate

More information

Hypergeometric Distributions

Hypergeometric Distributions 7.4 Hypergeometric Distributios Whe choosig the startig lie-up for a game, a coach obviously has to choose a differet player for each positio. Similarly, whe a uio elects delegates for a covetio or you

More information

CURIOUS MATHEMATICS FOR FUN AND JOY

CURIOUS MATHEMATICS FOR FUN AND JOY WHOPPING COOL MATH! CURIOUS MATHEMATICS FOR FUN AND JOY APRIL 1 PROMOTIONAL CORNER: Have you a evet, a workshop, a website, some materials you would like to share with the world? Let me kow! If the work

More information

Vladimir N. Burkov, Dmitri A. Novikov MODELS AND METHODS OF MULTIPROJECTS MANAGEMENT

Vladimir N. Burkov, Dmitri A. Novikov MODELS AND METHODS OF MULTIPROJECTS MANAGEMENT Keywords: project maagemet, resource allocatio, etwork plaig Vladimir N Burkov, Dmitri A Novikov MODELS AND METHODS OF MULTIPROJECTS MANAGEMENT The paper deals with the problems of resource allocatio betwee

More information

Measures of Spread and Boxplots Discrete Math, Section 9.4

Measures of Spread and Boxplots Discrete Math, Section 9.4 Measures of Spread ad Boxplots Discrete Math, Sectio 9.4 We start with a example: Example 1: Comparig Mea ad Media Compute the mea ad media of each data set: S 1 = {4, 6, 8, 10, 1, 14, 16} S = {4, 7, 9,

More information

Analysis Notes (only a draft, and the first one!)

Analysis Notes (only a draft, and the first one!) Aalysis Notes (oly a draft, ad the first oe!) Ali Nesi Mathematics Departmet Istabul Bilgi Uiversity Kuştepe Şişli Istabul Turkey [email protected] Jue 22, 2004 2 Cotets 1 Prelimiaries 9 1.1 Biary Operatio...........................

More information

Center, Spread, and Shape in Inference: Claims, Caveats, and Insights

Center, Spread, and Shape in Inference: Claims, Caveats, and Insights Ceter, Spread, ad Shape i Iferece: Claims, Caveats, ad Isights Dr. Nacy Pfeig (Uiversity of Pittsburgh) AMATYC November 2008 Prelimiary Activities 1. I would like to produce a iterval estimate for the

More information

Lecture 5: Span, linear independence, bases, and dimension

Lecture 5: Span, linear independence, bases, and dimension Lecture 5: Spa, liear idepedece, bases, ad dimesio Travis Schedler Thurs, Sep 23, 2010 (versio: 9/21 9:55 PM) 1 Motivatio Motivatio To uderstad what it meas that R has dimesio oe, R 2 dimesio 2, etc.;

More information

BENEFIT-COST ANALYSIS Financial and Economic Appraisal using Spreadsheets

BENEFIT-COST ANALYSIS Financial and Economic Appraisal using Spreadsheets BENEIT-CST ANALYSIS iacial ad Ecoomic Appraisal usig Spreadsheets Ch. 2: Ivestmet Appraisal - Priciples Harry Campbell & Richard Brow School of Ecoomics The Uiversity of Queeslad Review of basic cocepts

More information

CHAPTER 3 DIGITAL CODING OF SIGNALS

CHAPTER 3 DIGITAL CODING OF SIGNALS CHAPTER 3 DIGITAL CODING OF SIGNALS Computers are ofte used to automate the recordig of measuremets. The trasducers ad sigal coditioig circuits produce a voltage sigal that is proportioal to a quatity

More information

Designing Incentives for Online Question and Answer Forums

Designing Incentives for Online Question and Answer Forums Desigig Icetives for Olie Questio ad Aswer Forums Shaili Jai School of Egieerig ad Applied Scieces Harvard Uiversity Cambridge, MA 0238 USA [email protected] Yilig Che School of Egieerig ad Applied

More information

*The most important feature of MRP as compared with ordinary inventory control analysis is its time phasing feature.

*The most important feature of MRP as compared with ordinary inventory control analysis is its time phasing feature. Itegrated Productio ad Ivetory Cotrol System MRP ad MRP II Framework of Maufacturig System Ivetory cotrol, productio schedulig, capacity plaig ad fiacial ad busiess decisios i a productio system are iterrelated.

More information

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 8

CME 302: NUMERICAL LINEAR ALGEBRA FALL 2005/06 LECTURE 8 CME 30: NUMERICAL LINEAR ALGEBRA FALL 005/06 LECTURE 8 GENE H GOLUB 1 Positive Defiite Matrices A matrix A is positive defiite if x Ax > 0 for all ozero x A positive defiite matrix has real ad positive

More information

A Mathematical Perspective on Gambling

A Mathematical Perspective on Gambling A Mathematical Perspective o Gamblig Molly Maxwell Abstract. This paper presets some basic topics i probability ad statistics, icludig sample spaces, probabilistic evets, expectatios, the biomial ad ormal

More information

4.3. The Integral and Comparison Tests

4.3. The Integral and Comparison Tests 4.3. THE INTEGRAL AND COMPARISON TESTS 9 4.3. The Itegral ad Compariso Tests 4.3.. The Itegral Test. Suppose f is a cotiuous, positive, decreasig fuctio o [, ), ad let a = f(). The the covergece or divergece

More information