A PRESENTATION THEOREM FOR CONTINUOUS LOGIC AND METRIC ABSTRACT ELEMENTARY CLASSES

Size: px
Start display at page:

Download "A PRESENTATION THEOREM FOR CONTINUOUS LOGIC AND METRIC ABSTRACT ELEMENTARY CLASSES"

Transcription

1 A PRESENTATION THEOREM FOR CONTINUOUS LOGIC AND METRIC ABSTRACT ELEMENTARY CLASSES WILL BONEY Cotets 1. Itroductio 1 2. Models ad Theories 3 3. Elemetary Substructure Types ad Saturatio T dese as a Abstract Elemetary Class Metric Abstract Elemetary Classes 18 Refereces Itroductio I the spirit of Chag ad Shelah s presetatio results (from [Cha68] ad [Sh88], respectively), we prove a presetatio theorem for classes of cotiuous structures, both those axiomatized by first-order ad beyod, i terms of a class of discrete structures. The thrust of this presetatio theorem is the basic aalytic fact that the behavior of cotiuous fuctios is determied by their values o a dese subset of their domai. Focusig o dese subsets is key because it allows us to drop the requiremet that structures be complete, which is ot a property expressible by discrete (classical) logic, eve i the broader cotexts of L λ,ω or Abstract Elemetary Classes. The specific statemets of the presetatio theorems appear below (see Theorem 2.1 for cotiuous first-order logic ad Theorem 6.1 for Metric Abstract Elemetary Classes), but the geeral idea is the same i both cases: give a cotiuous laguage L, we defie a discrete laguage L + that allows us to approximate the values of the fuctios ad relatios by a coutable dese subset of values, amely Q [0, 1]. Note that the specificatio that this dese set (ad its completio) is stadard already requires a L + ω 1,ω setece, eve if we are workig i cotiuous first-order logic. The, give a cotiuous L-structure M ad a icely dese (see Defiitio 1.1 below) subset of it A, we ca form a discrete L + -structure A with uiverse A that ecodes all of M. Date: August 15,

2 2 WILL BONEY Give a class of cotiuous L-structures, we represet it by all dese approximatios (i the way described above) to members of the class. I each presetatio theorem, the way the cotiuous class is defied helps defie the class of approximatios. If the class is axiomatized by a cotiuous first-order theory, the the uiform cotiuity of the fuctios ad relatios (see Be-Yaacov, Berestei, Heso, ad Usvyatsov [BBHU08]) allow a L + ω 1,ω axiomatizatio of the dese approximatios. If the cotiuous class is o-elemetary ad forms a Metric Abstract Elemetary Class (see Hirvoe ad Hyttie [HH09] or Villaveces ad Zambrao [ViZa]), the the class of dese approximatios has o explicit axiomatizatio, but forms a Abstract Elemetary Class. We carry the relatio betwee the class of cotiuous models ad discrete approximatios further by cosiderig various model-theoretic properties ad fidig aalogues for them i the class of approximatios. We cosider types, saturatio, (ad for Metric Abstract Elemetary Classes), amalgamatio, joit embeddig, ad d-tameess. Throughout we assume that the reader is familiar with the basics of the cotiuous cotexts either first-order or Metric Abstract Elemetary Classes but try to provide refereces ad remiders whe discussig the cocepts. Parts of this paper come from the author s Ph.D. thesis uder the directio of Rami Grossberg at Caregie Mello Uiversity ad I would like to thak Professor Grossberg for his guidace ad assistace i my research i geeral ad i this work specifically. This material is based upo work while the author was supported by the Natioal Sciece Foudatio uder Grat No. DMS This paper represets a i progress draft. May of the argumets i the two cases are similar. I order to avoid repeatig the same argumets twice i slightly differet cotexts, we provide the details oly oce. We have chose to provide the details for cotiuous first-order logic because this cotext ofte allows more specific formulatios of the correspodece. Dese sets are ot quite the right cotext because they eed ot be substructures of the larger structure. Istead, we itroduce icely dese sets to require them to be closed uder fuctios. Defiitio 1.1. Give a cotiuous model M ad a set A M, we say that A is icely dese iff A is dese i the metric structure ( M, d M ) ad A is closed uder the fuctios of M. I the what follows, we will ofte wat to prove similar results for both greater tha ad less tha. I order to avoid writig everythig twice, we ofte use to stad i for both ad. Thus, assertig a statemet for r s meas that that statemet is true both for r s ad for r s. Our goal is to traslate the real-valued formulas of L ito classic, true/false formulas of L +. We do this by ecodig relatios ito L + that are iteded to specify the value of φ by decidig if it is above or below each possible value.

3 T dese AND K dese 3 To esure that the size of the laguage does t grow, we take advatage of the separability or R ad oly compare each φ to the ratioals i [0, 1]. For otatioal ease, we set Q := [0, 1] Q. 2. Models ad Theories The mai thesis of the presetatio of cotiuous first-order logic is that modeltheoretic properties of cotiuous first order structures ca be traslated to modeltheoretic (but typically quatifier free) properties of discrete structures that model a specific theory i a expaded laguage. We use F ml c L to deote the cotiuous formulas of the laguage L. The mai theorem about this presetatio is the followig: Theorem 2.1. Let L be a cotiuous laguage. The there is (a) a discrete laguage L + ; (b) a L + ω 1,ω theory T dese ; (c) a map that takes cotiuous L-structures M ad icely dese subsets A to discrete L + -structures M A that model T dese ; (d) a map that takes discrete L + structures A that model T dese to cotiuous L-structures A with the properties that (1) M A T dese has uiverse A ad, for ay a A, φ(x) F ml c L, r Q, ad stadig for ad, we have M A R φ r [a] φ M (a) r (2) A is a dese subset of A ad, for ay a A, φ(x) F ml c L, r Q, ad stadig for ad, we have A R φ r [a] φ A (a) r (3) these maps are (essetially) each other s iverse. That is, give ay icely dese A M, we have M = A M A ad, give ay L + -structure A T dese, we have (A) A = A. The essetially i the last clause comes from the fact that completios are ot techically uique as the objects selected as limits ca vary, but this fairly pedatic poit is the oly obstacle. Restrictig to dese subsets ad their completios have already bee cosidered i cotiuous first-order logic, where it goes by the ame prestructure (see [BBHU08]. 3). The key differece here is that, while prestructures are still cotiuous objects with uiformly cotiuous fuctios ad relatios, M A is a discrete object with the relatios o it beig either true or false. Proof: Our proof is log, but straightforward. First, we will defie L + ad T dese. The, we will itroduce the map (M, A) M A ad prove it satisfies (1). After this, we will itroduce the other map A A ad prove (2). Fially, we will

4 4 WILL BONEY prove that they satisfy (3). Defiig the ew laguage ad theory We defie the laguage L + to be F + i, R φ(x) r, R φ(x) r i<f,φ(x) F ml c (L),r Q with the arity of F + i matchig the arity of F i ad the arity of R φ(x) r matchig l(x). Sice we oly use a full (dese) set of coectives (see [BBHU08].6.1), we have esured that L + = L + ℵ 0. We defie T dese L + ω 1,ω to be the uiversal closure of all of the followig formulas ragig over all cotiuous formulas φ(z) ad ψ(z ), all terms τ(z, z ), ad all r, s Q ad t Q {0}. We have divided them ito headigs so that their meaig is (hopefully) more clear. Whe we refer to specific seteces of T dese later, we referece the orderig i this list. As always, a i a formula meas that it should be icluded with both a ad a replacig the. (1) The ordered structure of R (a) R φ(z) r (x) = R φ(z) r (x) (b) R φ(z) r (x) = R φ(z) r (x) (c) If r > s, the iclude R φ(z) r (x) R φ(z) s (x) (d) If r s, the iclude R φ(z) s (x) = R φ(z) r (x); ad R φ(z) r (x) = R φ(z) s (x) (e) R φ(z) r (x) R φ(z) r (x) (f) <ω r,s Q, r s < 1 R φ r (x) R φ s (x) (g) ( <ω R φ(z) r 1 (x)) = R φ(z) r (x) (h) ( <ω R φ(z) r+ 1 (x)) = R φ(z) r (x) (2) Costructio of formulas (a) R φ(z) 0 (x) R φ(z) 1 (x) (b) R 0 t (x) R 1 1 t (x); (c) R φ(z) r(x) R φ(z) 2r(x) 2 (d) R φ(z) r(x) R φ(z) 2r(x); 2 (e) R φ(z) ψ(z ) r(x, x ) s Q (R ψ(z ) s(x ) R φ(z) r+s (x)) (f) R φ(z) ψ(z ) r(x, x ) s Q ( R ψ(z ) s(x ) R φ(z) r+s (x)); (g) R supy τ(y,y) r(x) xr τ(y,y) r (x, x) (h) R supy τ(y,y) r(x) <ω xr τ(y,y) r 1 (i) R ify τ(y,y) r(x) <ω xr τ(y,y) r+ 1 (j) R ify τ(y,y) r(x) xr φ(y,y) r (x, x); (k) R φ(y,y) r (τ(x ), x) R φ(τ(y ),x) r(x, x) (3) Metric structure (a) R d(y,y ) 0(x, x ) x = x ; (x, x) ; (x, x)

5 T dese AND K dese 5 (b) R d(y,y ) r(x, x ) R d(y,y ) r(x, x); (c) r Q (R d(y,y ) r(x, x ) = x s Q [0,r]R d(y,y ) s(x, x ) R d(y,y ) r s(x, x )) (4) Uiform Cotiuity (a) For each r, s Q ad i < L F such that s < Fi (r), we iclude the setece i< R d(z,z ) s(x i, y i ) = R d(z,z ) r(f i (x), F i (y)) (b) For each r, s Q ad j < L R such that s < Rj (r), we iclude the setece i< R d(z,z ) s(x i, y i ) = (R Rj (z) R j (z ) r(x, y) R Rj (z) R j (z ) r(y, x)) We have bee careful about the specific eumeratio of these axioms for a reaso. If the origial cotiuous laguage is coutable, the T dese is coutable. I particular, we could take the cojuctio of it ad make it a sigle L + ω 1,ω setece. This meas that it is expressible i a coutable fragmet of L ω1,ω. Coutable fragmets are the most well-studied ifiitary laguages ad may of the results i, say, Keisler [Kei71] use these fragmets. I geeral, T dese is expressible i a L + ℵ 0 sized fragmet of L + ω 1,ω. From cotiuous to discrete... This is the easier of the directios. We defie the structure M A so that all of the iteded correspodeces hold ad everythig works out well. Suppose we have a cotiuous L-structure M ad a icely dese subset A. Now we defie a L + structure M A by (1) the uiverse of M A is A; (2) (F + i ) M A = Fi M A for i < F ; ad (3) for r Q ad φ(x) F ml c (L), set R M A φ r = {a A : φm (a) r} This is a L + -structure sice it is closed uder fuctios. The real meat of this part is the followig claim, which is (1) from the theorem. Claim: M A T dese ad, for ay a A, φ(x) F ml c L, r Q, ad =,, we have M A R φ r [a] φ M (a) r Proof of Claim: This is all straightforward. From the defiitio, we kow that, for ay a A ad formula φ(x) F ml c (L) ad {, }, we have M A R φ r [a] φ M (a) r

6 6 WILL BONEY This gives a easy proof of the fact that M A T dese because they are all just true facts if R φ r (a) is replaced by φ(a) r. Claim...ad back agai This is the harder directio. We wat to read out the L-structure that A is a dese subset of from the L + structure. First, we use the axioms of T dese to show that we ca read out the metric ad relatios of L from the relatios of L + ad that these are well-defied. The we complete A ad use the uiform cotiuity of the derived relatios to expad them to the whole structure. I the first directio, T dese could have bee ay collectio of true seteces about cotiuous structures ad the real lie, but this directio makes it clear that the axioms chose are sufficiet. Suppose that we have a L + -structure A that models T dese. The followig claim is a importat step i readig out the relatios of the completio of A from A. Claim 2.2. For ay φ(x) F ml c (L) ad a A, we have sup{t Q : A = R φ(x) t (a)} = if{t Q : A = R φ(x) t (a)} Proof: We show this equality by showig two iequalities. Let r {t Q : A = R φ(x) t (a)} ad s {t Q : A = R φ(x) t (a)}. The A = R φ(x) r (a) R φ(x) s (a) The, sice M + satisfies 1c, we must have r s. Thus sup{t Q : A = R φ(x) t (a)} if{t Q : A = R φ(x) t (a)}. By 1f, we have A = <ω r,s Q ; r s < 1 R φ(x) r (a) R φ(x) s (a) Let ɛ > 0. The there is 0 < ω such that ɛ > 1 0. By the above, there are r, s Q such that r s < 1 0 ad M + = R φ(x) r (a) R φ(x) s (a) As above, 1c implies r s, so we have r s < 1 0 < ɛ. Thus r < s + ɛ ad s {t Q : A = R φ(x) t (a)} ad r {t Q : A = R φ(x) t (a)}. The, if{t Q : A = R φ(x) t (a)} sup{t Q : A = R φ(x) t (a)}. Claim The first relatio that we eed is the metric. Give a, b A, we set D(a, b) := sup{r Q : A R d(x,y) r [a, b]} = if{r Q : A R d(x,y) r [a, b]}

7 T dese AND K dese 7 These defiitios are equivalet by Claim 2.2. We show that this is ideed a metric o A. Claim 2.3. ( A, D) is a metric space. Proof: We go through the metric space axioms. Let a, b A. (1) D(a, b) = 0 = if{r Q : A R d(x,y) r (a, b)} = 0 (2) = < ω r Q so A = R d(x,y) r (a, b) ad r 1 = 1d < ω, A = R d(x,y) 1 (a, b) = 1h A = R d(x,y) 0 (a, b) = a = b a = b = A = R d(x,y) 0 (a, b) = if{r Q : A = R d(x,y) r (a, b)} = 0 = D(a, b) = 0 D(a, b) = sup{r Q : A = R d(x,y) r (a, b)} = 3b sup{r Q : A = R d(x,y) r (b, a)} = D(b, a) (3) Let c A. We wat to show D(a, c) D(a, b) + D(b, c). It is eough to show r Q (D(a, c) r = D(a, b) + D(b, c) r) Thus, let r Q ad suppose D(a, c) r. The sup{s Q : A = R d(x,y) s (a, c)} r. By 3c, this meas sup{s Q : A = t Q [0,s]R d(x,y) t (a, b) R d(x,y) s t (b, c)} r Fix < ω. There is some s Q such that s r 1 ad A = t Q [0,s ]R d(x,y) t (a, b) R d(x,y) s t(b, c) Thus, there is some t Q such that 0 t s ad A = R d(x,y) t (a, b) R d(x,y) s t (b, c) By the defiitio of D, this meas that D(a, b) t ad D(b, c) s t ; thus, D(a, b) + D(b, c) s. Sice this is true for all < ω, we get that D(a, b) + D(b, c) r as desired. Thus, D is a metric o A. Claim Now we defie partial fuctios ad relatios o ( A, D) such that they are uiformly cotiuous. I particular, (1) for i < L F, set f i := Fi A with modulus fi (r) = sup{s Q : A x 0,..., x (Fi ) 1; y 0,..., y (Fi ) 1 ( i<(fi )R d(z,z ) s(x i, y i ) R d(z,z ) r(f i (x), F i (y))}.

8 8 WILL BONEY (2) for j < L R, set r j (a) := sup{r Q : A R Rj (z) r[a]} with modulus rj (r) = sup{s Q : A = x y( i<(rj )R d(z,z ) s(x i, y i ) (R Rj (z) R j (z ) r(x, y) R Rj (z) R j (z ) r(y, x)))}. These fuctios are ot defied o the desired structure (ie the completio of A), but they already fulfill our goal i terms of agreeig with the discrete relatios i the followig sese. Claim: For all a A ad all formulas φ(x) built up from these fuctios ad D, we have that φ(a) r A R φ(z) r [a] Proof: We proceed by iductio o the costructio of φ(x). We assume that is i our proofs, but the proofs for are the same. If φ is atomic, the it falls ito oe of the followig cases. Suppose φ(x) R j (τ(x)) for some term τ. The R M j (τ(a)) r if{s Q : A = R Rj (x) s[τ(a)]} r 1d 1g 2k < ω, s Q so s r 1 ad A = R R j (x) s [τ(a)] < ω, A = R Rj (x) r 1 [τ(a)] A = R Rj (x) r[τ(a)] A = R Rj (τ(y)) r[a] Suppose that φ(x, y) d(τ 1 (x), τ 2 (y)) for terms τ 1 ad τ 2. The detail are essetially as above: D M (τ 1 (a), τ 2 (b)) iff (by 1d, the defiitio of sup, ad 1h ad 1g) M + = R d(x,y) r [τ 1 (a), τ 2 (b)] iff (by 2k) M + = R d(τ1 (x),τ 2 (y)) r(a, b). For the iductive step, we deal with each coective (from our full set) i tur. The iductio steps for x 0, x 1, ad x x are clear. 2 Suppose φ ψ τ, where τ is a formula ad ot a term. Note if r = 0, the this is obvious. So assume r 0. Recall that φ M (a) = ψ M (a) τ M (a) = { ψ M (a) τ M (a) if ψ M (a) 0 0 otherwise Thus, we ca assume we are i the case that ψ M (a) > τ M (a). First, suppose ψ M (a) τ M (a) r. Sice ψ M (a) > τ M (a), there is some s Q such that ψ M (a) > s > τ M (a). The τ M (a) s ad ψ M (a) s + r. By iductio, we have that M + = R τ(x) s [a] R ψ M (x) s+r[a] The, by 2e, we have that M + = R ψ τ r[a] as desired.

9 T dese AND K dese 9 Now, suppose M + = R ψ τ r[a]. Agai, 2e implies there is is some s Q such that M + = R τ s [a] R ψ r+s [a] By iductio, we get τ M (a) s ad ψ M (a) r + s. The φ M (a) = ψ M (a) τ M (a) (r + s) s = r as desired. Suppose φ(x) sup x ψ(x, x). We will cosider both sides of the iequality sice they re ot symmetrically axiomatized (see 2g ad 2h), but we wo t worry about if. Suppose that sup x φ M (x, a) r. The for ay < ω, there is some a M such that φ M (a, a) > r 1 2. Sice φm is uiformly cotiuous, there is some δ > 0 such that, if d(a, b) < δ, the φ M (a, a) φ M (b, a) < 1 2. Sice M + is dese i M, there is some a M + such that d(a, a ) < δ. Thus, φ M (a, a) > r 1. By iductio, we have that M + = <ω xr φ(y,y) r 1 (x, a) The 2h says that M + = R supy φ(y,y) r(a). Suppose that M + = R supy φ(y,y) r[a]. The, by 2h, M + = <ω xr φ(y,y) r 1 (x, a). So, for each < ω, there is some a M + such that M + = R φ(y,y) r 1 [a, a]. By iductio, we have that φ M (a, a) r 1. Sice this is true for each < ω, we get sup y φ M (y, a) r. The other directio is easier ad we ca combie the two parts sup φ M (x, a) r a Mφ M (a, a) r x a M + φ(a, a) r Iductio M + = xr φ r (x, a) 2g M + = R supx φ(x,x)[a] We have give these fuctios moduli, but do ot kow they are uiformly cotiuous. We show this ow. It is also worth otig that these moduli might ot be the same moduli i the origial sigature L. Istead, these are the optimal moduli, while the origial laguage might have moduli that could be improved. Claim: The fuctios f i ad r j are cotiuous. Proof: We do each of these cases separately.

10 10 WILL BONEY Sub-Claim 1: F M + i is uiformly cotiuous o ( M +, D) with modulus Fi. Let r Q ad let a, b M + such that max i< D(a i, b i ) < Fi (r). Thus, for each i <, D(a i, b i ) = if{s Q : M + = R d(x,y) s (a i, b i )} < Fi (r). Sice this is strict, there is some s i Q such that M + = R d(x,y) si (a i, b i ). Note that 1d implies that the set Fi (r) is supremumig over is dowward closed. Thus, s = max i< s i is i it. Thus, we ca coclude M + = R d(x,y) r [F i (a), F i (b)] This meas that D(F i (a), F i (b)) r, as desired. Sub-Claim 2: R M + j is uiformly cotiuous o ([0, 1], ) with modulus Rj. Let r Q ad a, b M + such that i< D(a i, b i ) < Rj (r). From the ifimum defiitio of D, for each i <, there is s i Q such that s i < Rj (r) ad M + = R d(z,z ) s i [a i, b i ]. Thus, M + = R Rj (z) R j (z ) r(a, b) R Rj (z) R j (z ) r(b, a) For this ext part, we eed some of the future proofs, but essetially we have eough to show that this implies R M + j (a) R M + j (b) r ad R M + j (b) R M + j (a) r This implies R M + j (a) R M + j (b) r, so R M + j is uiformly cotiuous. Now we have a prestructure, see [BBHU08]. 3. Now we complete A to A i the stadard way; see Mukries [Mu00] for a referece for the topological facts. I particular, we defie the cotiuous L structure A by A is the completio of ( A, D); the metric d A is the extesio of D to A ; for i < F, Fi A is the uique extesio of f i to A ; ad for j < R, Rj A is the uique extesio of r j to A. Essetial iverses Propositio 2.4. Give ay cotiuous L-structure M ad dese subset A, we have that M = A M A ad, give ay L + structure A that models T dese, we have that (A) A = A. Proof: First, let M be a cotiuous L-structure ad A M be icely dese. We defie a map f : M (M A ) as follows: if a A, the f(a) = a. For a M A, fix some (ay) sequece a A : < ω such that lim a = a (this limit computed i M). We kow that a : < ω is Cauchy i M, so it s

11 T dese AND K dese 11 Cauchy i (M A ). The set f(a) = lim a, where that limit is computed i (M A ). This is well-defied ad a bijectio because A is dese i both sets. That this is a L-isomorphism follows from applyig the correspodece twice: for all a A ad φ(x) F ml c (L) φ M (a) r M A R φ(x) r [a] φ (M A) (a) r ad the fact that the values of φ o A determies its values o M ad (M A ). Secod, let A be a L + structure that models T dese. Clearly, the uiverses are the same, ie, (A) A = A. For ay relatio R φ r ad a A, we have Give a fuctio F + i A R φ r [a] φ A (a) r (A) A R φ r [a] ad a, a A, we have that (F + i ) A (a) = a A R d(f + i (x),x) 0[a, a] + (A) A R d(f + i (x),x) 0[a, a] (F i ) (A) A (a) = a We ca exted this correspodece to theories. Suppose that T is a cotiuous theory i L. Followig [BBHU08].4.1, theories are sets of closed L-coditios; that is, a set of φ = 0, where φ is a formula with o free variables. The followig is immediate from Theorem 2.1. Corollary 2.5. If φ = 0 is a closed L-coditio, the φ M = 0 M A R φ 0 With our fixed theory T, set T to be T dese {R φ 0 : φ = 0 T }. The our represetatio of cotiuous L-structures as discrete L + -structures modelig T dese ca be exteded to a represetatio of cotiuous models of T ad discrete models of T. 3. Elemetary Substructure We ow discuss traslatig the otio of elemetary substructure betwee our two cotexts. Depedig o the geerality eeded, this is either easy or difficult. For the easy case, we have the followig. Theorem 3.1. Let M, N be cotiuous L structures. The M L N iff, for every icely dese A M ad B N such that A B, we have that M A L + N B. Note that the relatio betwee M A ad N B is just substructure. So eve though they are models of ifiitary theories, their relatio just cocers atomic formulas. This is because we have built the quatifiers of L ito the relatios of L +. Proof: : Let A = M ad B = N. The M N, so M M L + N N by assumptio. Thus they agree o all relatios cocerig elemets of M. Now we

12 12 WILL BONEY wat to show that M c L N. Let φ(x) F mlc L ad a M. From the theorems proved last sectio, we have, for each r Q, φ M (a) r M M R φ(x) r [a] Theorem 2.1 N N R φ(x) r [a] M M L + N N φ N (a) r Theorem 2.1 Thus φ M (a) = φ N (a) ad M L N as desired. : Let A M ad B N be icely dese so A B. We wat to show that M A L + N B. Let F + L + ad a A. The, by defiitio of the structures, (F + ) M A (a) = F M (a) = F N (a) = (F + ) N B (a) Let R φ r (x) L + ad a A. M M R φ(x) r [a] φ M (a) r Theorem 2.1 Similarly, we have the followig. φ N (a) r M N N N R φ(x) r [a] Theorem 2.1 Theorem 3.2. Give A, B T dese, if A L + B, the A L B. However, these are ot the best theorems possible. I particular, the requiremet that A B limits the scope of this theorem. We would like to kow whe L + structures complete to L-elemetary substructures eve whe the dese substructures are ot subsets or each other; for istace, the completios of Q [0, 1] ad Q + 2 are icely related, but the previous theorem does ot see that. We would like to develop a criterio for L + structures A, B T dese that is equivalet to A L B. Our first attempt is the followig. Theorem 3.3. Suppose M N are cotiuous L-structures ad A M ad B N are icely dese. The (1) M A L + N C, where C is the closure of A B uder the fuctios of N. (2) There is a icely dese B N such that M A L + N B ad B = A + dc(n) + L. Proof: (1) Note that A C ad C is icely dese i N. By Theorem 3.1, M A L + N C. (2) Let B N be dese of size dc(n) ad let B be the closure of B A uder the fuctios of N. By Theorem 3.1, M A L + N B.

13 T dese AND K dese 13 While this is a improvemet, it is still ot the best desireable. I particular, it still makes referece to the cotiuous structures. We would prefer a correspodece that oly ivolved L + structures. To that ed, we give the followig defiitio of iessetial extesios. Defiitio 3.4. Give A L + B, we say that B is a iessetial extesio of A A iff for every b B ad < ω, there is some a A such that B R d(x,y)< 1 [b, a]. Propositio 3.5. If B is a iessetial extesio of A, the A = B. This gives us the followig theorem. Theorem 3.6. Let A, B T dese. The TFAE (1) A L B. (2) There is a L + structure C such that A, B L + C ad C is a iessetial A extesio of B. (3) There is a L + structure C T dese such that A, B L + C ad C is a iessetial extesio of B. (4) There is a extesio of the fuctios ad relatios of L + to A B such that A B T dese that are still uiformly cotiuous ad such that for every a A B ad < ω, there is some b B such that A B R d(x,y)< 1 [b, a]. Proof: (1) = (2) Take C = (B) A B. (2) = (3) Immediate. (3) = (4) Take the extesio iherited from C. (4) = (1) We have A, B A B from the first coditio ad B = A B form the secod. 4. Types ad Saturatio I this sectio, we will coect types i the cotiuous logic sese to types i the discrete sese. However, just as elemets i the complete structure are represeted by sequeces of the discrete structure, we represet types by sequece types. Recall from [BBHU08].8.1 that a type over B is a cosistet collectio of coditios of the form φ(x, b) = 0 with b B. Defiitio 4.1. We say that r : < ω is a sequece l-type over B iff r 0 (x) is a l-type over B ad r +1 (x, y) is a 2l-type over B such that there is some idex set I, (possibly repeatig) formulas φ i : i I ; ad (possibly repeatig) Cauchy sequeces b i B <ω : i I so d(b i, b i +1) 1 2 such that r 0 (x) = {R φi (z,z ) w i(2)(x, b i 0) : i I}; ad φ

14 14 WILL BONEY r +1 (x, y) = {R φi (z,z ) w φ i( 1 2 ) (x, bi +1) : i I} {R d(z,z ) 1 2 (x k, y k ) : k < l} A realizatio of a sequece type r : < ω is a : < ω such that a 0 realizes r 0 ; ad a +1 a realizes r +1. Note that the use of 1 is ot ecessary; this could be replaced by ay summable 2 1 sequece for a equivalet defiitio (also replacig by the trailig sums). 2 1 However, we fix 1 for computatioal ease. The fudametal coectio betwee 2 cotiuous types ad sequece types is the followig. Theorem 4.2. Let A M be icely dese. (1) If B M ad r(x) is a partial l-type over B, the for ay B A such that B B, there is a sequece l type r : < ω over B such that M realizes r iff M A realizes r : < ω (2) If B A ad r : < ω is a partial sequece l-type over B, the there is a uique l-type r over B such that M realizes r iff M A realizes r : < ω We ca deote the type i (2) by lim r. I each case, we have that a M A : < ω realizes r : < ω implies lim a realizes lim r. Proof: (1) Recall that r(x) cotais coditios of the form φ(x, b) = 0 for φ F ml c (L) ad b B. For < ω ad b B, set B (b) = {b B : d M (b, b) < 1 }. To make the cardiality work out icer, fix a choice 2 fuctio G, ie G(B (b)) B (b). The B (b) ad G(B (b)) have the obvious meaigs. Defie r + (x) := {R φ(z;y)<w φ 1 ( (b))) : φ(x; b) = 0 r} 2 1 r 0 (x) := r 0 + (x) r +1 (x, y) := r + +1(x) {R d(z,z )< 1 2 (x i, y i ) : i < l(x)} The r : < ω is a sequece type over B ; we ca see this by takig r as the idex set, φ i = φ, ad b i = G(B (b)) for i = φ(x; b) = 0 r. To show it has the desired property, first suppose that a : < ω from M A realizes r : < ω. We kow that a : < ω is a Cauchy sequece; i particular, for m >, d M (a, a m ) m i= 1 2 = 2m+1 1 i 2 m+1

15 T dese AND K dese 15 Sice M A = M is complete, there is a MA such that lim a = a. We claim that a r. Let φ(x; b) = 0 r. The d M A (ab; a G(B (b))) = max{d M (a, a ), d M (b, G(B (b))} 1 max{ 2, 1 i 2 } = Thus, φ M (a; b) φ M (a ; G(B (b))) < w φ 1 ( 2 ) 1 Lettig, we have that φ M A (a; b) = lim φ M (a ; G(B (b))) lim w φ 1 ( 2 ) = 0 1 as desired. Now suppose that a M realizes r. Sice A is dese, we ca fid a A : < ω such that d M (a, a +1 ) < 1 ad a 2 a. We wat to show that r + (a ) holds. Let φ(x, b) r. We kow that d M (a G(B (b)), ab) = 1 2 1, so we get φ M A (a, G(B (b))) = φ M (a, b) φ M (a, G(B (b))) < w φ 1 ( 2 ) 1 as desired. (2) Let r : < ω be a partial sequece l-type give by I, φ i : i I, ad b i <ω : i I. The set r(x) := {φ i (x, lim b i ) = 0 : i I} First, suppose that a M A : < ω realizes r : < ω. The, sice a +1 a r +1, we have d M A (a +1, a ) < 1 ad, thus, the sequece is 2 Cauchy. Sice M is complete, let a = lim a M. The, by uiform cotiuity, we have φ M i (a, b i ) = φ M i ( lim a, lim b i ) = lim φ M i (a, b i ) lim w φ i 1 ( 2 ) 1 = 0 So a r. Now suppose that a M realizes t. The, by deseess, we ca fid a Cauchy sequece a A : < ω such that d(a +1, a ) 1. The 2 d(ab i, a b i ) 1. The we ca coclude 2 1 i=

16 16 WILL BONEY φ M i (a, b i ) φ M i (a, b i ) w φ i 1 ( 2 ) 1 φ M A (a, a i ) w φ i 1 ( 2 ) 1 So a : < ω realizes r : < ω. We ow coect type-theoretic cocepts i cotiuous logic (e.g. saturatio ad stability) with cocepts i our discrete aalogue. Recall (see [BBHU08].7.5) that a cotiuous structure M is κ-saturated iff, for ay A M of size < κ ad ay cotiuous type r(x) over A, if every fiite subset of r(x) is satisfiable i M, the so is r(x). Defiitio 4.3. If r : < ω is a sequece type defied by a idex set I ad I 0 I, the r : < ω I 0 is the sequece type defied by I 0. We say that M A T dese is κ-saturated for sequece types iff, for all B A ad sequece type r : < ω over B that is defied by I, if r : < ω I 0 is realized i M A for all fiite I 0 I, the r : < ω is realized i M A. Theorem 4.4. (1) If M is κ-saturated ad λ ℵ 0 < κ, the M A is λ + saturated for sequece types. (2) If M A is κ saturated, the M is κ saturated. Proof: (1) Let M be κ-saturated ad A M be icely dese. Let B M A of size λ ad let r : < ω be a sequece type over B that is fiitely satisfiable i M A. Set r = lim r from Theorem 4.2; this is a type over B where B λ ℵ 0 < κ. We claim that r is fiitely satisfiable i M. Ay fiite subset of r of r = {φ i (x, lim b i ) : i I} correspods to a fiite I 0 I. The, by Theorem 4.2, r 0 is realized i M iff r : < ω I 0 is realized i M A. The, sice each r : < ω I 0 is realized i M A by assumptio, we have that r is fiitely satisfiable i M. By the κ-saturatio of M, r is realized i M. By Theorem 4.2, r : < ω is realized i M A. So M A is λ + -saturated. (2) Let M A be κ-saturated for sequece types. Let B M of size < κ ad r be a type over B that is fiitely satisfied i B. Fid B A such that B B; this ca be doe with B B + ℵ 0 < κ. The form the sequece type r : < ω over B that coverges to r as i Theorem 4.2. As before, sice r is fiitely satisfiable i M, so is r : < ω i M A. So r : < ω is realized i M A by saturatio. Thus, r is realized i M. We immediately get the followig corollary.

17 T dese AND K dese 17 Corollary 4.5. If κ = (λ ℵ 0 ) + or, more geerally, κ = sup λ<κ (λ ℵ 0 ) + ad M is of size κ, the M is saturated iff M A is saturated for some icely dese A M of size κ. 5. T dese as a Abstract Elemetary Class I this sectio we view the discrete side of thigs as a Abstract Elemetary Class; see Baldwi [Bal09] or Grossberg [Gro1X]. Theorem 5.1. Let T be a complete, cotiuous first order L-theory. The let L + ad T dese be from Theorem 2.1. Set K = (Mod (T dese T, L +). The (1) K is a AEC; (2) K has amalgamatio, joit embeddig, ad o maximal models; ad (3) Galois types i K correspod to sequece types (Defiitio 4.1). Note that if T were ot complete, the amalgamatio would ot hold. However, the other properties will cotiue to hold, icludig the correspodece betwee Galois types ad sequece types. Proof: T dese T + is a L ω1,ω theory, so all of the examples hold except perhaps the chai axioms. For those, cosider a L +-icreasig chai M Ai : i < α. The, by Theorem 2.1 ad Theorem 3.2, the sequece M Ai : i < α is L -icreasig chai that each model T. The by the chai axiom for cotiuous logic, there is M = i<α (M Ai ) that models T. Additioally, A := i<α A i is icely dese i M. Thus, M A = i<α M Ai is as desired. Additioally, if M Ai L + M B for some B, the M L M B, so M A L + M B. These properties all follow from the correspodig properties of cotiuous firstorder logic. For istace, cosiderig amalgamatio, suppose M A L + M B, M C. The we have M A L M B, M C. By amalgamatio for cotiuous first-order logic, there is some N L M B ad elemetary f : M C M A N. Let D N be icely dese that cotais B C. The we have M B L + N D ad f M C : M C MA N D ; this is a amalgamatio of the origial system. Fially, we wish to show that Galois types are sequece types ad vice versa. Note that there are moster models i each class. Further more, we may assume that, if C is the moster model of T, that there is some icely dese U C such that the moster model of K is M U ; i fact, we could take U = C. Let cotiuous M T ad A M be icely dese. If we have tuples a ad b, the gtp K (a/m A ) = gtp K (b/m A ) f Aut MA M U.f(a) = b f Aut MA C.f(a) = b tp(a/m) = tp(b/m) lim r a = lim r b

18 18 WILL BONEY where r x : < ω is the sequece type derived from tp(x/m) as i Theorem 4.2 usig A as the dese subset. 6. Metric Abstract Elemetary Classes I this sectio, we exted the above represetatio to Metric Abstract Elemetary Classes. Recall from Hirvoe ad Hyttie [HH09] or that a Metric Abstract Elemetary Class (MAEC) is a class of cotiuous L-structures K ad a strog substructure relatio K satisfyig the followig axioms: (1) K is a partial order o K; (2) for every M, N K, if M K N, the M L N; (3) (K, K ) respects L(K) isomorphisms, if f : N N is a L(K) isomorphism ad N K, the N K ad if we also have M K with M K N, the f(m) K ad f(m) K N ; (4) (Coherece) if M 0, M 1, M 2 K with M 0 K M 2 ; M 1 K M 2 ; ad M 0 M 1, the M 0 M 1 ; (5) (Tarski-Vaught chai axioms) suppose M i K : i < α is a K -icreasig cotiuous chai, the (a) i<α M i K ad, for all i < α, we have M i K i<α M i ; ad (b) if there is some N K such that, for all i < α, we have M i K N, the we also have i<α M i K N; ad (6) (Löweheim-Skolem umber) There is a ifiite cardial λ L(K) such that for ay M K ad A M, there is some N K M such that A N ad N A + λ. We deote the miimum such cardial by LS(K). These axioms were first give i Hirvoe ad Hyttie [HH09]. A key differece is that the fuctios ad relatios L are o loger required to be uiformly cotiuous, but just cotiuous. This is due to the lack of compactess i the MAEC cotext. This iitially seems problematic because fuctios must be uiformly cotiuous o a set to be guarateed a extesio to its closure. However, we get aroud this by simply defiig K dese to be all the structures that happe to complete to a member of K, the use the MAEC axioms to show that K dese satisfies the AEC axioms. For this reaso, whe we refer to cotiuous laguages, structures, etc. i this sectio, we will ot mea that they are uiformly cotiuous. Theorem 6.1. Let L be a cotiuous laguage. The there is a discrete laguage L + such that, for every MAEC K with LS(K) = L, there is (1) a AEC K dese with L(K dese ) = L + ad LS(K dese ) = LS(K); (2) a map from M K ad icely dese subsets A of M to M A K dese ; ad (3) a map from A K dese to A K with the properties that

19 T dese AND K dese 19 (1) M A has uiverse A ad for each a A, r Q, ad {, }, we have that M A R Rj (z) r[a] R M j (a) r (2) A has uiverse that is the completio of A with respect to the derived metric ad for each a A, r Q, ad {, }, we have that R A j (a) r A R Rj (z) r[a] (3) The maps above are essetially iverses, i the sice of Theorem (4) Give M l K ad A l icely dese i M l for l = 0, 1, if f : M 0 M 1 is a K-embeddig such that f(a 0 ) A 1, the f A 0 is a K dese - embeddig from (M 0 ) A0 to (M 1 ) A1. Give A, B K dese ad a K dese -embeddig f : A B, this lifts caoically to a K-embeddig f : A B. Proof: The proof proceeds similar to the first-order versio, Theorem 2.1. I particulalr, may of the defiitios of cotiuous structures from discrete approximatios (such as gettig the metric ad relatios from their approximatios) did ot use compactess ad oly used uiform cotiuity to esure that a completio existed, which will be guarateed by the defiitio of K dese i this case. Give cotiuous L = F i, R j i<f,j< r, defie L + := F + i, R Rj (z) r, R Rj (z) r i<f,j< r,r Q Give a MAEC K, we defie the AEC K dese as follows: L(K dese ) = L + ; Give a L + structure A, we use the followig procedure to determie membership i K dese : defie D o A A by D(a, b) := sup{r Q : A R d(x,y) r [a, b]} = if{r Q : A R d(x,y) r [a, b]} The proof that this is a well-defied ad is a metric proceeds exactly as i the previous case. We ca similarly defie the relatios R j o A ad complete the uiverse (A, D) to A. We call the structure A completable iff (1) for every a A ad every Cauchy sequece a A : < ω covergig to to a, the value of lim R j (a ) is idepedet of the choice of the sequece; ad similarly (2) for every a A ad every Cauchy sequece a A : < ω covergig to to a, the value of lim F + i (a ) is idepedet of the choice of the sequece. If A is completable, the we defie A to be the L + -structure where F i ad R j are defied o A accordig to the idepedet value give above. Fially, we say that A K dese iff (1) A is completable; ad

20 20 WILL BONEY (2) A K. Give A, B K dese, we say that A dese B iff (1) A L + B; ad (2) A K B. Now that we have the defiitio of K dese, we must show that it is i fact a AEC. The verificatio of the axioms are routie; we give the argumets for coherece ad the chai axioms as templates. For coherece, suppose that A, B, C K dese such that A dese C; B dese C; ad A L + B. The, takig completios, we get that A K C; B K C; ad A L B By coherece i K, we the have that A K B. The, by defiitio, A dese B, as desired. For the chai axioms, suppose that A i K dese : i < α is a cotiuous, dese -icreasig chai such that, for all i < α, A i dese B K dese. Agai, takig completios, we get that A i K : i < α is a cotiuous, K -icreasig chai such that, for all i < α, A i K B K. The, by the uio axioms for K, we have that i<α A i K ad i<α A i K B. Note that the existece of i<α A i shows that i<α A i is completable ad that a easy computatio shows that i<α A i = i<α A i. Thus, i<α A i K dese ad, for all j < α, A j dese i<α A i dese B Oce we have defied the maps ad show that K dese is a AEC, the rest of the proof proceeds exactly as i the cotiuous first-order case, i some ways simpler sice L + oly has relatios for each relatio of L, rather tha each formula of L. We ow tur to a applicatio. Both Hirvoe ad Zambrao have proved versios of Shelah s Presetatio Theorem for MAECs i their theses. The more geeral is Zambrao s Theorem from [Zam]: Theorem 6.2. Let K be a MAEC. There is L 1 L ad a cotiuous 1 L 1 -theory T 1 ad a set of T 1 -types Γ such that K = P C(T 1, Γ, L). A immediate corollary to our presetatio theorem is a discrete presetatio theorem. Corollary 6.3 (Discrete Presetatio Theorem for Metric AECs). Let K be a MAEC. The there is a (discrete) laguage L 1 of size LS(K), a L 1 -theory T 1, ad a set of T 1 -types Γ such that K = {M 1 L(K) : M 1 T 1 ad omits Γ}, where the completio is take with respect to a caoically defiable metric. Proof: Apply Theorem 6.1 to represet K as a discrete AEC K dese ad the apply Shelah s Presetatio Theorem from Shelah [Sh88]. 1 But ot uiformly cotiuous

21 T dese AND K dese 21 Additioally, Zambrao asks (as Questio [Zam].1.2.9) if there exists is a Haf umber for model existece i MAECs. Usig our presetatio theorem, we ca aswer this questios i the affirmative. Furthermore, the Haf umber for MAECs is the same as for AECs. Theorem 6.4. If K is a MAEC with LS(K) = κ models of size or desity character cofial i ℶ (2 κ ) +, the K has models with desity character arbitrarily large. Proof: For every, λ < ℶ (2 κ ) +, let M λ K have size λ. M λ is icely dese i itself, so (M λ ) Mλ K dese has size λ. By the defiitio of Haf umber for discrete AECs, this meas that K dese has arbitrarily large models. Takig completios, this meas K has arbitrarily large models. The proof for desity character is the same. Give this represetatio, we ca determie basic structural properties of K by lookig at K dese ad vice versa. The above theorem already shows how to trasfer arbitrary large models ad the other properties trasfer similarly. Propositio 6.5. Suppose K is a MAEC. For P beig amalgamatio, joit embeddig, or o maximal models, K has P iff K dese has P. The fourth clause of the coclusio of Theorem 6.1 is stated as it is to make the proof of this propositio easy to see. We ow look at the otio of type i K dese that correspods to Galois types i K. As before, we pass to a sequece of types represetig a Cauchy sequece for a realizatio of the Galois type we wish to represet. Defiitio 6.6. Give A K dese, r : < ω is a sequece Galois type over A iff (1) r 0 gs l (A); ad (2) r +1 gs l2 (A) such that (a) if ab r +1, the d(a, b) 1 ; ad 2 (b) r {l,...,l2 1} +1 = r, l i. e. the first l coordiates of r are the same as the fial l coordiates of r +1. Give a sequece Galois type r : < ω over A ad A dese B, we say that a B : < ω realizes r : < ω iff (1) a 0 r 0 ; ad (2) a +1 a r +1 Give a sequece Galois type r : < ω over A ad A dese B, we say that a B : < ω weakly realizes r : < ω iff there is some C ad b C : < ω such that (1) B dese C; (2) b : < ω realizes r : < ω ; ad (3) lim <ω a = lim <ω b.

22 22 WILL BONEY Note that realizig a sequece Galois type is differet tha realizig each idividual Galois type i the sequece separately. However, we ca always realize sequece Galois types give amalgamatio. Lemma 6.7. Suppose that K has amalgamatio. Give ay sequece Galois type r : < ω over A K dese, there is A dese B K dese that cotais a realizatio of r : < ω. Proof: By the defiitio of Galois type, we ca write each r as gtp Kdese (a 0 0/A; B 0 ) ad gtp Kdese (a +1 +1a +1 /A; B +1 ). Usig amalgamatio i K dese, which follows from amalgamatio i K, we ca costruct icreasig C : < ω ad icreasig f : B A C such that f 0 is the idetity ad f (a ) = f +1 (a +1 ); this secod part is due to the secod clause i the defiitio of sequece types. Settig C = <ω C, we have that f (a ) : < ω realizes r : < ω. Ufortuately, we do t have the same tight coectio betwee Galois types i K ad sequece Galois types i K dese as exists i Theorem 4.2. This is due to the fact that the Galois versio of sequece types specifies a distace betwee cosecutive members of the Cauchy sequece, rather tha just specifyig a boud o the distace. It is possible that perturbatios (as i Hirvoe ad Hyttie [HH12]) might be used to restore this coectio. Istead, we have itroduced the otio of weakly realizig a sequece Galois type because this is eough to prove a variat of Theorem 4.4 i this cotext. Theorem 6.8. Let M K ad A K dese such that A = M. (1) If M is κ-galois saturated ad λ ℵ 0 < κ, the A is λ + -weakly saturated for sequece Galois types (i.e. give ay A 0 dese A of size < λ +, every sequece Galois type over A 0 is weakly realized i A). (2) If A is κ-weakly saturated for sequece Galois types, the M is κ-galois saturated. Proof: First, suppose that M is κ saturated ad λ ℵ 0 < κ. Let A 0 dese A be of size λ ad let r : < ω be a sequece Galois type over A 0. By Lemma 6.7, there is some B dese A 0 ad a : < ω that realizes r : < ω. After completig the members of K dese, we have A 0 M ad A 0 B K. Sice a B : < ω is a Cauchy sequece, there is some a B such that lim a = a. Sice M is κ-saturated ad A 0 λ ℵ 0 < κ, there is b M that realizes gtp K (a/a 0 ; B). Sice A is dese i M, there is some Cauchy sequece b A : < ω that coverges to b. The b : < ω weakly realizes r : < ω. Secod, suppose that A is κ-weakly saturated for sequece Galois types ad let M 0 M of size < κ ad r gs(m 0 ). Let A 0 A be icely dese i M 0 ; the A 0 := (M 0 ) A0 dese A. I K, we ca write r as gtp K (a/m 0 ; N). The, i K dese, r := gtp Kdese (a/a 0 ; N N ) : < ω is a sequece Galois type over A 0. Sice A 0 < κ, by A s weak saturatio, there is some b : < ω that weakly realizes

23 T dese AND K dese 23 r. This meas that we ca fid a extesio B dese A ad f : N N A0 B such that lim f(a) = lim b Sice b A, this meas that f(a) M. Sice N N is complete, the K dese - embeddig f is i fact a K-embeddig from N ito B ad fixes A 0 = M 0. Thus, we have that f(a) M realizes gtp K (a/m 0 ; N). A crucial property i the study of MAECs is whether the atural otio of distace betwee Galois types defies a metric or ot. This property is called the Pertubatio Property by Hirvoe ad Hyttie [HH09] ad the Cotiuity Type Property by Zambrao [Zam12]. For ease, we assume that K (equivaletly, K dese ) has a moster model C. Defiitio 6.9. Give M K ad p, q S(M), we defie d(p, q) = if{d(a, b) : a p ad b q} K has the Pertubatio Property (PP) iff, give ay Cauchy sequece b C : < ω ad M K, if, for all < m < ω, gtp(b /M) = gtp(b m /M) the, gtp(lim b /M) = gtp(b 0 /M). Although these properties might ot iitially seem related, a little work shows that d is always a pseudometric ad that it is a metric iff PP holds; this is due Hirvoe ad Hyttie. The, similar to tameess from AECs, Zambrao [Zam12].2.9 defies a otio of tameess i MAECs that satisfy PP. Defiitio K is µ-d-tame iff for every ɛ > 0, there is a δ > 0 such that for every M K of desity character µ ad p, q gs(m), if d(p, q) ɛ, the there is some N M of desity character µ such that d(p N, q N) δ ɛ. Agai these properties trasfer to sequece types i K dese. We ca defie a pseudometric o sequece Galois types i K dese by d dese ( r : < ω, s : < ω ) := if{ lim d(a, b ) : a : < ω realizes r : < ω, b : < ω realizes s : < ω } The d dese is a metric iff d is, ad µ-d-tameess trasfers from K to K dese i the obvious way. Refereces [Bal09] Joh Baldwi, Categoricity, Uiversity Lecture Series, America Mathematical Society, [BBHU08] Itai Be Yaacov, Alexader Berestei, C. Ward Heso, Alexader Usvyatsov, Model theory for metric structures, Model theory with applicatios to algebra ad aalysis, vol. 2 (Chatzidakis, Macpherso, Pillay, Wilke eds.). Lodo Math Society Lecture Note Series, Cambridge Uiversity Press, 2008.

24 24 WILL BONEY [Cha68] C. C. Chag, Some remarks o the model theory of ifiitary laguages, The Sytax ad Sematics of Ifiitary Laguages (J. Barwise, ed.), vol. 72, [Gro1X] Rami Grossberg, A Course i Model Theory, I Preparatio, 201X. [HH09] Åsa Hirvoe ad Tapai Hyttie, Categoricity i homogeeous complete metric spaces, Archive of Mathematical Logic 48 (2009), o. 3-4, [HH12] Åsa Hirvoe ad Tapai Hyttie, Metric Abstract Elemetary Classes with pertubatios, Fudameta Mathematica 217 (2012), [Kei71] H. Jerome Keisler, Model Theory for Ifiitary Logic, Studies i Logic ad the Foudatios of Mathematics, vol. 62, North-Hollad Publishig Co., Amsterdam, [Mu00] James Mukres, Topology, 2d ed., Pretice Hall 2000 [Sh88] Saharo Shelah, Classificatio of oelemetary classes, II. Abstract elemetary classes, Classificatio theory (Joh Baldwi, ed.), 1987, pp [ViZa] Adres Villaveces ad Pedro Zambrao, Aroud superstability i Metric Abstract Elemetary Classes: limit models ad r-towers, Accepted, Preprit series of Mittag-Leffler Istitute. [Zam] Pedro Zambrao, Aroud superstability i Metric Abstract Elemetary Classes, Ph. D thesis, [Zam12] Pedro Zambrao, A stability trasfer theorem for d-tame Metric Abstract Elemetary Classes, Math Logic Q (58), p , address: wboey2@uic.edu Mathematics, Statistics, ad Computer Sciece Departmet, Uiversity of Illiois at Chicago, Chicago, Illiois, USA

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

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

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

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

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

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

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

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

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

Irreducible polynomials with consecutive zero coefficients

Irreducible polynomials with consecutive zero coefficients Irreducible polyomials with cosecutive zero coefficiets Theodoulos Garefalakis Departmet of Mathematics, Uiversity of Crete, 71409 Heraklio, Greece Abstract Let q be a prime power. We cosider the problem

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

Lecture 13. Lecturer: Jonathan Kelner Scribe: Jonathan Pines (2009)

Lecture 13. Lecturer: Jonathan Kelner Scribe: Jonathan Pines (2009) 18.409 A Algorithmist s Toolkit October 27, 2009 Lecture 13 Lecturer: Joatha Keler Scribe: Joatha Pies (2009) 1 Outlie Last time, we proved the Bru-Mikowski iequality for boxes. Today we ll go over the

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 faf@cam.ac.uk dot@dotat.at Uiversity of Cambridge Computig Service February 009 Abstract I these otes I eplai how to derive formulae for umerically

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

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

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

Infinite Sequences and Series

Infinite Sequences and Series CHAPTER 4 Ifiite Sequeces ad Series 4.1. Sequeces A sequece is a ifiite ordered list of umbers, for example the sequece of odd positive itegers: 1, 3, 5, 7, 9, 11, 13, 15, 17, 19, 21, 23, 25, 27, 29...

More information

5 Boolean Decision Trees (February 11)

5 Boolean Decision Trees (February 11) 5 Boolea Decisio Trees (February 11) 5.1 Graph Coectivity Suppose we are give a udirected graph G, represeted as a boolea adjacecy matrix = (a ij ), where a ij = 1 if ad oly if vertices i ad j are coected

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

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

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

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

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

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

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 aesi@bilgi.edu.tr Jue 22, 2004 2 Cotets 1 Prelimiaries 9 1.1 Biary Operatio...........................

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

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

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

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

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

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

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

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

THIN SEQUENCES AND THE GRAM MATRIX PAMELA GORKIN, JOHN E. MCCARTHY, SANDRA POTT, AND BRETT D. WICK

THIN SEQUENCES AND THE GRAM MATRIX PAMELA GORKIN, JOHN E. MCCARTHY, SANDRA POTT, AND BRETT D. WICK THIN SEQUENCES AND THE GRAM MATRIX PAMELA GORKIN, JOHN E MCCARTHY, SANDRA POTT, AND BRETT D WICK Abstract We provide a ew proof of Volberg s Theorem characterizig thi iterpolatig sequeces as those for

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

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

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

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

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

Perfect Packing Theorems and the Average-Case Behavior of Optimal and Online Bin Packing

Perfect Packing Theorems and the Average-Case Behavior of Optimal and Online Bin Packing SIAM REVIEW Vol. 44, No. 1, pp. 95 108 c 2002 Society for Idustrial ad Applied Mathematics Perfect Packig Theorems ad the Average-Case Behavior of Optimal ad Olie Bi Packig E. G. Coffma, Jr. C. Courcoubetis

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

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

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

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

Metric, Normed, and Topological Spaces

Metric, Normed, and Topological Spaces Chapter 13 Metric, Normed, ad Topological Spaces A metric space is a set X that has a otio of the distace d(x, y) betwee every pair of poits x, y X. A fudametal example is R with the absolute-value metric

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

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 iva.kojadiovic@uiv-ates.fr Jea-Luc Marichal Applied Mathematics

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

On the L p -conjecture for locally compact groups

On the L p -conjecture for locally compact groups Arch. Math. 89 (2007), 237 242 c 2007 Birkhäuser Verlag Basel/Switzerlad 0003/889X/030237-6, ublished olie 2007-08-0 DOI 0.007/s0003-007-993-x Archiv der Mathematik O the L -cojecture for locally comact

More information

LECTURE 13: Cross-validation

LECTURE 13: Cross-validation LECTURE 3: Cross-validatio Resampli methods Cross Validatio Bootstrap Bias ad variace estimatio with the Bootstrap Three-way data partitioi Itroductio to Patter Aalysis Ricardo Gutierrez-Osua Texas A&M

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

.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

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

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

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

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

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

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

Lecture 3. denote the orthogonal complement of S k. Then. 1 x S k. n. 2 x T Ax = ( ) λ x. with x = 1, we have. i = λ k x 2 = λ k.

Lecture 3. denote the orthogonal complement of S k. Then. 1 x S k. n. 2 x T Ax = ( ) λ x. with x = 1, we have. i = λ k x 2 = λ k. 18.409 A Algorithmist s Toolkit September 17, 009 Lecture 3 Lecturer: Joatha Keler Scribe: Adre Wibisoo 1 Outlie Today s lecture covers three mai parts: Courat-Fischer formula ad Rayleigh quotiets The

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

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

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

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

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

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

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

CONTINUOUS AND RANDOM VAPNIK-CHERVONENKIS CLASSES

CONTINUOUS AND RANDOM VAPNIK-CHERVONENKIS CLASSES CONTINUOUS AND RANDOM VAPNIK-CHERVONENKIS CLASSES ITAÏ BEN YAACOV Abstract. Nous démotros que si T est ue théorie dépedate, sa radomisée de Keisler T R l est aussi. Pour faire cela ous gééralisos la otio

More information

A Faster Clause-Shortening Algorithm for SAT with No Restriction on Clause Length

A Faster Clause-Shortening Algorithm for SAT with No Restriction on Clause Length Joural o Satisfiability, Boolea Modelig ad Computatio 1 2005) 49-60 A Faster Clause-Shorteig Algorithm for SAT with No Restrictio o Clause Legth Evgey Datsi Alexader Wolpert Departmet of Computer Sciece

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

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

A Recursive Formula for Moments of a Binomial Distribution

A Recursive Formula for Moments of a Binomial Distribution A Recursive Formula for Momets of a Biomial Distributio Árpád Béyi beyi@mathumassedu, Uiversity of Massachusetts, Amherst, MA 01003 ad Saverio M Maago smmaago@psavymil Naval Postgraduate School, Moterey,

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

Research Article Sign Data Derivative Recovery

Research Article Sign Data Derivative Recovery Iteratioal Scholarly Research Network ISRN Applied Mathematics Volume 0, Article ID 63070, 7 pages doi:0.540/0/63070 Research Article Sig Data Derivative Recovery L. M. Housto, G. A. Glass, ad A. D. Dymikov

More information

Lesson 17 Pearson s Correlation Coefficient

Lesson 17 Pearson s Correlation Coefficient Outlie Measures of Relatioships Pearso s Correlatio Coefficiet (r) -types of data -scatter plots -measure of directio -measure of stregth Computatio -covariatio of X ad Y -uique variatio i X ad Y -measurig

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

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

Lecture 4: Cheeger s Inequality

Lecture 4: Cheeger s Inequality Spectral Graph Theory ad Applicatios WS 0/0 Lecture 4: Cheeger s Iequality Lecturer: Thomas Sauerwald & He Su Statemet of Cheeger s Iequality I this lecture we assume for simplicity that G is a d-regular

More information

Annuities Under Random Rates of Interest II By Abraham Zaks. Technion I.I.T. Haifa ISRAEL and Haifa University Haifa ISRAEL.

Annuities Under Random Rates of Interest II By Abraham Zaks. Technion I.I.T. Haifa ISRAEL and Haifa University Haifa ISRAEL. Auities Uder Radom Rates of Iterest II By Abraham Zas Techio I.I.T. Haifa ISRAEL ad Haifa Uiversity Haifa ISRAEL Departmet of Mathematics, Techio - Israel Istitute of Techology, 3000, Haifa, Israel I memory

More information

Exploratory Data Analysis

Exploratory Data Analysis 1 Exploratory Data Aalysis Exploratory data aalysis is ofte the rst step i a statistical aalysis, for it helps uderstadig the mai features of the particular sample that a aalyst is usig. Itelliget descriptios

More information

G r a d e. 2 M a t h e M a t i c s. statistics and Probability

G r a d e. 2 M a t h e M a t i c s. statistics and Probability G r a d e 2 M a t h e M a t i c s statistics ad Probability Grade 2: Statistics (Data Aalysis) (2.SP.1, 2.SP.2) edurig uderstadigs: data ca be collected ad orgaized i a variety of ways. data ca be used

More information

Project Deliverables. CS 361, Lecture 28. Outline. Project Deliverables. Administrative. Project Comments

Project Deliverables. CS 361, Lecture 28. Outline. Project Deliverables. Administrative. Project Comments Project Deliverables CS 361, Lecture 28 Jared Saia Uiversity of New Mexico Each Group should tur i oe group project cosistig of: About 6-12 pages of text (ca be loger with appedix) 6-12 figures (please

More information

Plug-in martingales for testing exchangeability on-line

Plug-in martingales for testing exchangeability on-line Plug-i martigales for testig exchageability o-lie Valetia Fedorova, Alex Gammerma, Ilia Nouretdiov, ad Vladimir Vovk Computer Learig Research Cetre Royal Holloway, Uiversity of Lodo, UK {valetia,ilia,alex,vovk}@cs.rhul.ac.uk

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

A RANDOM PERMUTATION MODEL ARISING IN CHEMISTRY

A RANDOM PERMUTATION MODEL ARISING IN CHEMISTRY J. Appl. Prob. 45, 060 070 2008 Prited i Eglad Applied Probability Trust 2008 A RANDOM PERMUTATION MODEL ARISING IN CHEMISTRY MARK BROWN, The City College of New York EROL A. PEKÖZ, Bosto Uiversity SHELDON

More information

Determining the sample size

Determining the sample size Determiig the sample size Oe of the most commo questios ay statisticia gets asked is How large a sample size do I eed? Researchers are ofte surprised to fid out that the aswer depeds o a umber of factors

More information

Lesson 15 ANOVA (analysis of variance)

Lesson 15 ANOVA (analysis of variance) Outlie Variability -betwee group variability -withi group variability -total variability -F-ratio Computatio -sums of squares (betwee/withi/total -degrees of freedom (betwee/withi/total -mea square (betwee/withi

More information

PART TWO. Measure, Integration, and Differentiation

PART TWO. Measure, Integration, and Differentiation PART TWO Measure, Itegratio, ad Differetiatio Émile Félix-Édouard-Justi Borel (1871 1956 Émile Borel was bor at Sait-Affrique, Frace, o Jauary 7, 1871, the third child of Hooré Borel, a Protestat miister,

More information

Factoring x n 1: cyclotomic and Aurifeuillian polynomials Paul Garrett <garrett@math.umn.edu>

Factoring x n 1: cyclotomic and Aurifeuillian polynomials Paul Garrett <garrett@math.umn.edu> (March 16, 004) Factorig x 1: cyclotomic ad Aurifeuillia polyomials Paul Garrett Polyomials of the form x 1, x 3 1, x 4 1 have at least oe systematic factorizatio x 1 = (x 1)(x 1

More information

The following example will help us understand The Sampling Distribution of the Mean. C1 C2 C3 C4 C5 50 miles 84 miles 38 miles 120 miles 48 miles

The following example will help us understand The Sampling Distribution of the Mean. C1 C2 C3 C4 C5 50 miles 84 miles 38 miles 120 miles 48 miles The followig eample will help us uderstad The Samplig Distributio of the Mea Review: The populatio is the etire collectio of all idividuals or objects of iterest The sample is the portio of the populatio

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

ON THE DENSE TRAJECTORY OF LASOTA EQUATION

ON THE DENSE TRAJECTORY OF LASOTA EQUATION UNIVERSITATIS IAGELLONICAE ACTA MATHEMATICA, FASCICULUS XLIII 2005 ON THE DENSE TRAJECTORY OF LASOTA EQUATION by Atoi Leo Dawidowicz ad Najemedi Haribash Abstract. I preseted paper the dese trajectory

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

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

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

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

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

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

Subject CT5 Contingencies Core Technical Syllabus

Subject CT5 Contingencies Core Technical Syllabus Subject CT5 Cotigecies Core Techical Syllabus for the 2015 exams 1 Jue 2014 Aim The aim of the Cotigecies subject is to provide a groudig i the mathematical techiques which ca be used to model ad value

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

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