Hacia un Modelo de Red Inmunológica Artificial Basado en Kernels. Towards a Kernel Based Model for Artificial Immune Networks
|
|
- Cordelia Melina Leonard
- 8 years ago
- Views:
Transcription
1 Haca un Modeo de Red Inmunoógca Artfca Basado en Kernes Towards a Kerne Based Mode for Artfca Immune Networs Juan C. Gaeano, Ing. 1, Fabo A. Gonzáez, PhD. 1 Integent Systems Research Lab, Natona Unversty of Coomba Branch Bogota Bongenum Research Lab, Natona Unversty of Coomba Branch Bogota {cgaeanoh, fagonzaezo}@una.edu.co Recbdo para revsón 8 de Novembre de 007, aceptado 14 de Febrero de 008, versón fna 8 de Febrero de 008 Resumen En este artícuo se presenta una adaptacón de un modeo de red nmunoógca artfca a a estratega de os métodos de erne. Esta adaptacón proporcona a área de os sstemas nmunoógcos artfcaes, por prmera vez, agunas de as ventaas de os métodos de erne, taes como a habdad para ser apcados sobre datos no vectoraes y a transformacón a espacos de ata dmensonadad por medo de amado erne trc. Agunos exper mentos premnares fueron evados a cabo con e fn de obtener agunas uces sobre e comportamento de modeo propuesto. Paabras Cave sstemas ntegentes, agr upamento basado en smtud, redes nmunoógcas artfcaes, métodos de er ne. Abstract Ths paper presents an adaptaton to the strategy of erne meth ods of a we n own ar tfca mmune n etwor. Ths adaptaton brngs to artfca mmune systems, for the frst tme, some of the advantages of erne methods, such as the abty to dea wth non vector data and the mappng to hgh dmensona spaces thr ough the erne tr c. Pr emnar y exper ments were carred out n order to get some nsghts of the behavor of the proposed mode. Keywords ntegent syst ems, smar ty based cust er n g, artfca mmune networs, er ne methods. I. INTRO DUCTION achne earnng oos for fndng and recognzng patterns M n data comng from a source of nterest []. The patterns extracted from such data may be used for obtanng nowedge of the data source or to mae predctons []. Machne earnng technques can be cassfed nto, at east, two approaches: the feature based approach and the smartybased approach [13]. The former consders a process to represent the obects as feature vectors as a prevous step to earnng []; the atter, uses ony a (ds)smarty measure between obects as the nput for the earnng process [13]. There are some domans where the obects have a compex structure, whch maes dffcut to perform the feature extracton process [13]. Exampes of ths nd of domans are mages, sequences, graphs, and text documents. In those cases, the appcaton of the smarty based approach appears to be more sutabe. Kerne methods s a machne earnng strategy that maes agorthm desgn ndependent from the representaton of the data [5]. Such ndependence requres that agorthms receve as nput a par wse measure nstead of feature vectors. From ths pont of vew we can consder the strategy of erne methods as a smarty based approach. Bo nspred computng (aso nown as natura computng) s the fed of machne earnng (and computer scence) that deveops probem sovng technques by mmcng a natura system. Some exampes of ths nd of technques are Genetc Agorthms that tae nspraton from speces evouton [9], Artfca Neura Networs that tae nspraton from bran functon [4], and Artfca Immune Systems (AIS) that tae nspraton from the mmune system behavor [6]. Most of those agorthms foow the feature based approach snce the frst step n settng them up s a representaton process where doman obects are usuay coded as attrbute vectors, wth each component representng some mportant feature of the obect. There s not much research on the deveopment of bo nspred agorthms that foow the smarty based approach, except for the erne versons of the sef organzng map (erne SOM) [16], [1], [1] and the erne partce swarm optmzaton (erne PSO) [1]. In ths paper, the probem of expressng an artfca mmune networ (AIN) mode as a erne method s taced. Specfcay, a nown AIN mode, anet [7], s modfed n such a way that t does not receve as nput the vector representaton of antgens, but the nter antgen dstances. Revsta Avances en Sstemas e Informátca, Vo.5 No. 1, Edcón Especa, Medeín, Mayo de 008, ISSN III Congreso Coombano de Computacón 3CCC 008
2 66 Revsta Avances en Sstemas e Informátca, Vo.5 No. 1, Edcón Especa, Medeín, Mayo de 008, ISSN III Congreso Coombano de Computacón 3CCC 008 Ths paper s organzed as foows: Secton II presents a revew of the artfca mmune networ mode for custerng and the strategy of erne methods for pattern anayss. Secton III presents an AIN mode for custerng based on the erne methods strategy. Secton IV presents the resuts of some premnary experments, and fnay, Secton V concudes the paper and proposes some deas for future wor. II. BACK GRO UND A. Artfca Immune Networs The natura mmune system (NIS) s the man defense of the body aganst pathogen organsms, caed antgens []. The NIS exhbts some nterestng propertes from a computatona pont of vew, such as pattern recognton, autonomy, dstrbutvty, earnng and memory [6]. Thus the fed of artfca mmune systems (AIS) wors on the defnton of modes nspred by prncpes and theores of the NIS n order to be apped to probem sovng [6]. The mmune earnng s the abty of the NIS for adaptng tsef to the changes n the antgenc envronment [6]. When an antgen enters the body for the frst tme, the NIS uses adaptaton mechansms to create ces (B ces) and moecues (antbodes) abe to dea wth such antgen. Addtonay, the NIS uses a memory mechansm for eepng an nterna mage of the antgen, so that t can dea wth t faster and more effectvey f the same antgen (or a smar one) nvades the body agan []. From the pont of vew of the mmune networ theory, mmune earnng and memory can be expaned as foows: the producton of a gven antbody (ected by an externa antgen) stmuates/ suppresses the producton of other antbodes that stmuate/ suppress the producton of other antbodes and so on []. The man hypothess of ths theory states that mmune memory s mantaned by B ces nteractng wth each other even n the absence of foregn antgens []. These nteractons between ces/moecues are va shape matchng (recognton). The way an antgen (or an antbody) can stmuate/suppress the producton of antbodes s based on the strength of the recognton, whch s caed affnty []. An artfca mmune networ (AIN) s an AIS that uses concepts from the mmune networ theory [6], many the nteractons between B ces (stmuaton and suppresson) and the ce conng and mutaton processes [8]. In the ast years dfferent AIN modes have been proposed to sove probems such as pattern recognton n DNA sequences [10], data anayss [0], [15], [7], [6], [7], web mnng [18], [19], mutmoda functon optmzaton [5], robot contro [11], [17], autonomous navgaton [14], [17] and e ma cassfcaton [4]. Most of AIN modes are based n the shape space concept ntroduced by Pereson [8]. Ths concept represents antbodes and antgens as ponts n a n dmensona space, where each dmenson s reated to a feature nvoved n the recognton process []. An antbody recognzes those antgens n ts scope, whch s defned by a hypersphere wth certan recognton radus and center n the antbody []. From ths pont of vew, AIN can be abeed as a feature based approach. Most of AIN modes assume a feature based representaton of the doman obects to be used n the earnng process. Ths maes dffcut to appy such modes n domans wth compex structure. Then, a representaton ndependent AIN mode woud aow to extend the appcaton doman of AIN. The erne methods approach gves a sutabe strategy to acheve such mode. B. Kerne Methods Kerne methods are descrbed as a unfed approach based on statstca methods for pattern anayss [5]. The approach assumes that transformng the nput data to a proper feature space coud mae the pattern dentfcaton process easer [5]. Such transformaton corresponds to the representaton process mentoned above. However, the transformaton s defned mpcty by a erne functon, whch s reated to the specfc data type and the nd of patterns that s expected to fnd n the data set [5]. Fg. 1. A schematc representaton of the erne methods approach The erne methods strategy can be represented as n Fg. 1. Data are mapped to a feature space through a erne functon κ, whch computes dot products between obects n the new space. The erne matrx Kcontans the evauaton of the erne functon for each par of obects and acts as nput of a earnng agorthm Athat s desgned to fnd patterns n the feature space based on such matrx [5]. One nterestng and usefu property of the erne functon s that t can compute the dot product from the orgna obects. Ths property s caed the erne trc [3]. Then, the earnng agorthm s desgned so that t does not need the coordnates of the obects n the new space, but ony the erne matrx [5]. Kerne methods can be apped to any data type,.e., nput data do not need to be represented as feature vectors [5]. It s because constrants to consder a functon as a erne functon are stated n terms of the erne matrx. Therefore, any par wse measure (e.g. smarty) abe to produce a par wse matrx that
3 Haca un Modeo de Red Inmunoógca Artfca Basado en Kernes Gaeano, Gonzáez 67 satsfes those constrants can be consdered as a erne functon [5]. Hence, erne methods can be abeed as a smarty based approach. Kerne methods have the advantage of separatng the earnng agorthm from the data representaton [5]. Ths aows research on pattern anayss to be defned n two drectons: one for defnng erne functons for specfc data types, and other for budng earnng agorthms to fnd specfc pattern types. The defnton of a erne method mpes the expresson of computatons n terms of dot products [5]. Ths can be acheved drecty as the case of support vector machnes or by modfyng a nown mode as the case of the erne means, whch s the erne verson of the means agorthm [5]. C. Kerne Methods on Bo nspred Agorthms There s not much research on the deveopment of bo nspred agorthms that foow the erne method strategy. Here, we revew the erne versons of the sef organzng map (SOM) and the erne verson of the partce swarm optmzaton agorthm (PSO). 1) KM Kerne SOM: In [16] MacDonad and Fyfe present a erne verson of the Kohonen's sef organzng map (SOM). The authors revew the erne means (KM), whch s the base for ths ernezed SOM (KSOM). In that mode, each mean m u s a near combnaton of nput data n the feature space, whch s expressed as x, where { } [ 1 N ] m µ γ µ ( x ) s the set of nput sampes, s the mappng from the nput space to the feature space, nduced by a erne functon, and γ μ s the weght of the th smpe for the mean m μ. Coeffcents γ μ are n [0, 1] and the sum over a the coeffcents representng a mean equas the number of ponts assgned to the custer that mean represents. Thus, n KM KSOM, neuron weghts are near combnatons of nput data n the feature space, and the weght updatng mechansm s the same as n the orgna SOM, even usng the same neghborhood functon. As revewed n [1], n ths mode neuron weghts are defned n the feature space but not n the nput space. ) GD Kerne SOM: Ths erne verson of SOM, revewed n [1], s presented n [1] by Pan et a. and n [3] by Andras. The mode s based on gradent descent (GD) and the probem s to mnmze the mean square error of an nput and the correspondng prototype n the mapped space. 3) EF Kerne SOM: In [1] Lau et a. present a revew of the KM KSOM the GD KSOM. They show that SOM can be performed competey n feature spaces and that both versons of KSOM are equvaent to a unfyng KSOM based on energy functons, a of ths eepng n mnd cassfcaton tass. Authors derve a KSOM based on a energy functon (EF) to be mnmzed. It s foowed from a defnton of the orgna SOM as mnmzer of such functon. They then show the updatng rues of KM SOM and GD SOM are partcuar cases of the updatng rue n the EF SOM, that can be used dependng on weghts are or are not defned n the feature space. 4) Kerne PSO: In [1] Abraham et a. present a erne verson of the Partce Swarm Optmzaton (PSO) agorthm to perform automatc custerng,.e., wthout pror nowedge about the rght number of custers. The modfed verson s acheved by defnng a custerng ndex, caed CS measure, n terms of nner products and sovng the custerng probem as an optmzaton probem where CS measure shoud be mnmzed, whch mpes an optma partton has been reached. In the process, each partce represents a partton of the nput data set and therefore, the souton s the gobay best partce at the end of the process. III. A PROPOSAL FOR A KERNEL ARTIFICIAL IMMUNE NETWORK (KAIN) Fg. shows a genera tranng agorthm for AIN, whch s taen from [8]. The agorthm receves as nput a set A A of antgens, whch are gong to be presented to the networ, where A s the set of a possbe antgens. The agorthm returns an artfca mmune networ composed of a set B B of B ces and connectons between them, where B s the set of a possbe B ces. The set B s the set of B ces of an AIN at a gven tme. The frst step, ntazaton, s to create an nta set of B ces. After ths, an teratve process s performed startng by presentng the set of antgens to the networ, whch means that for each antgen and each B ce the stmuaton s cacuated ( f A B A R ). Such stmuaton s based on an stmuato n : affnty measure between antgens and B ces ( f affnty : B A B A R ) that measures the smarty/ compementarty between eements n the shape space. In the next step B ces are aowed to nteract wth each other, ths s done by cacuatng the stmuaton and suppresson effects between them ( f B B B R stmuato n : and f B : B B sup presson R ). Smar to antgen/b ce stmuaton, B ce/b ce stmuaton (and suppresson) coud be cacuated as a functon of B ce/b ce affnty. Tota stmuaton of B ces s then cacuated by summng up the effects caused by antgen and networ nteractons ( F : B R ). Based on tota stmuaton, some B ces are seected and some copes ( f conng : B Ν ) of each seected B ce are created and mutated ( mutate : B B ). Some modes nterpret ths rate as the probabty of a B ce to be seected for sufferng mutaton; other modes nterpret t as the
4 68 Revsta Avances en Sstemas e Informátca, Vo.5 No. 1, Edcón Especa, Medeín, Mayo de 008, ISSN III Congreso Coombano de Computacón 3CCC 008 Fg.. A Genera Artfca Immune Networ agorthm proporton of the B ce feds that w be changed. In the meta dynamcs step, some useess B ces are removed from the networ, new B ces are created randomy and ncorporated nto the networ, and ns among a B ces are reorganzed. Fnay, when the stoppng crteron s met, the current networ s returned. A. A Modfed Verson of the GAIN Agorthm To modfy the GAIN agorthm so that t foows the strategy of erne methods, t shoud be notced that a cruca step s the mutaton operator, snce t s usuay mpemented as a few varaton n the coordnates of a feature vector. For ths purpose we propose to defne a new representaton for B ces. 1) Representaton and Mutaton of B Ces: Here, we foow a smar approach to that of Kerne SOM. Generay, n most AIN modes, antgens correspond to nput sampes and antbodes may be assmated to custer centrods (neura weghts). Therefore, n KAIN B ces (antbodes) are represented as near combnatons of antgens n the feature space. In a more forma way, a B ce s represented as b ( a ) where b B s a B ce, a A s an antgen, s a γ mappng from the nput space to a feature space and s a weght. We can constran the near combnaton to be convex n order to mae the B ce be surrounded by the antgens t recognzes. In genera, a may be any antgen presented to the networ,.e. A A, however we can restrct A to ncude ony those antgens whch have nteracted wth (stmuated) b. Notce that those A are not necessary dsont. From the boogca pont of vew, ths representaton maes sense because when an antgen enters the body, the mmune system creates some B ces n order to neutraze such antgen. The B ces more ey to survve are those wth hgh affnty wth the stmuatng antgen, n that way, those B ces are nterna mages of such antgen. As a B ce s stmuated by those antgens cose enough n the shape space,.e., the affnty between the B ce and each of those stmuatng antgens s greater than a certan threshod, the shape of a B ce s defned for the set of antgens t recognzes, as more than one B ce can recognze the same antgen, the sets A are not necessary dsont. A a, a B Notce that gven a set of antgens { } [ 1, m ] ce b s totay determned by the coeffcents γ. Ths
5 Haca un Modeo de Red Inmunoógca Artfca Basado en Kernes Gaeano, Gonzáez 69 means that a B ce b can be represented as the vector of weghts (γ 1..., γ m ). Based on ths representaton t s easy to adapt a genetc agorthm mutaton scheme for rea vaued chromosomes. For nstance, a Gaussan mutaton scheme as descrbed n [9] mutate b ( ) mutate ( γ 1,..., γ m ) ( γ,..., γ ) + (,..., ) 1 m where each Δ ~ N(0, σ) and σ s a parameter that contros the expected sze of Δ. Notce that a normazaton step s requred after addng the vector Δ n order to eep the convexty property. The mpementaton of ths operator s straghtforward, however the cruca step s the mpementaton of mutaton n such a representaton. ) B ce antgen and B ce B ce affnty: As t was mentoned n Secton II, affnty s a centra concept n AIN modes. Snce B ces are represented as a near combnaton of antgens, the B ce antgen and the B ce B ce affnty may be defned n terms of antgen antgen affnty. Notce that n the natura mmune system there s not such a concept as antgen antgen affnty; ths s aso the case n most AIN. However, most of the AIN represent B ces and antgens n the same way, thus the B ce antgen affnty measure may be naturay extended to an antgen antgen affnty measure. For ths paper, we use the antgen antgen affnty gven by f affnty A where ( ) ( ) A ( ( ) ( ) a, exp a D a a Gven that D A 1 m D σ x x, x, the expresson above turns nto ( a ) ( a ), ( a ) ( a ) ( a ), ( a ) + ( a ), ( a ) ( a ), ( a ) whch can be expressed usng the erne functon A ( a, a ) + κ ( a, a ) κ ( a a ) D κ, κ as In a smar way, the B ce antgen affnty w gven by where D BA a f affnty D σ BA ( ( )) a b, exp a γ b ( a ) κ ( a, a ) + κ ( a, a ) κ ( a, a ) + and fnay, the B ce B ce affnty can be expressed as where f D affnty B B D σ ( ) b, b exp b, b + and usng the fact that b, b b b b b, b b the expresson above turns nto D B b, b b, b ( a ), γ ( a ) γ κ γ γ κ γ ( a ), ( a ) κ ( a, a ) ( a, a ) ( a, a ) κ + + ( a, a ) 3) New B ce generaton: As the representaton of B ces n feature space requres a set of antgens for each B ce, the ntazaton step can be performed by randomy seectng a set A 0 Awhere each antgen represents a B ce,.e., f the th antgen s seected, the correspondng B ce w be b ( ) a and 1 for. γ where γ 0 for a IV. PRELIMINARY EXPERIMENTS Some smpe experments were carred out n order to test the sutabty of the proposed mode for detectng custers n data. In ths wor resuts w be vadated by vsua evauaton of B ces ocatons. For ths purpose a synthetc dmensona data set consstng of 5 custers was used. Ths data set s a verson of the one used n [7], t conssts of 50 ponts, each custer contanng 10 ponts. Fg. 3 shows a pot of ths data set. The mpementaton we used s an adaptaton of the anet agorthm [7] consderng the defntons presented n Secton III. One maor modfcaton was ntroduced n the affnty measure, whe anet agorthm uses 1 D as the affnty measure, where D represents the Eucdean dstance between patterns, the proposed agorthm uses D exp as the affnty measure. σ +
6 70 Revsta Avances en Sstemas e Informátca, Vo.5 No. 1, Edcón Especa, Medeín, Mayo de 008, ISSN III Congreso Coombano de Computacón 3CCC 008 A. Expermenta Setup In ths setup experments were carred out varyng the nput data: the frst one usng custers and 3 (ponts n the upper eft hand and ower rght hand), and the second one usng the whoe data set. Tabe I shows an expanaton of each parameter aong wth the vaues used n ths setup. For these experments, the dentty erne κ ( x, y ) x, y was used, whch means that no transformaton of the nput space s made. Ths has the purpose of the vsua vadaton of the agorthm's performance. B. Expermenta Resuts Fg. 4 and Fg. 5 show the output of the agorthm for the experments performed. Antgens are represented as crces and B ces are represented as crosses. Notce that n both experments, B ces are ocated n the antgen regons, and regons free from antgens reman empty. Ths suggests that the agorthm s abe detect custers. Another property that can be seen s that the number of B ces s ow compared to the number of antgens. Notce that ths agorthm does not perform data compresson, as the orgna anet does, because each antgen presented to the networ s ncuded n a near combnaton that defnes a B ce. Tang nto account these resuts, we can say that t performs we n spaces where custers appear to be we separated. As the mode has been presented as a erne method, we coud tae advantages from the erne trc n order to transform spaces of nput data wth dfferent custer structures nto spaces where they appear to be we separated. V. CO NCL USI O N In ths paper a erne based mode for artfca mmune networs was ntroduced. A ey concept for ths mode s the representaton of antbodes as near combnatons of antgens, whch aows the defnton of a mutaton mechansm wthout assumng a vector representaton of antbodes. The mode assumes an affnty measure between antgens, whch s an dea present n the current modes and supported by the shapespace concept, but not expcty mentoned. Expermenta resuts on synthetc data gave a good nsght of the abty of the proposed mode for detectng custers. Some ssues were mentoned regardng the modfed verson of the anet mode, whch suggest some tass for future wor, namey to test the use of the erne trc wth dfferent nds of custer structures, to ncude the dea of "antbody saturaton", whch means that a B ce can be defned by a mted number of antgens, and to appy the mode to a non vector data set. ACKNOWLEDGMENT Authors than to members of the Integent Research Lab. and Bongenum Research Lab. for ther support n ths proect. REFERENCES [1] Ath Abraham, Swagatam Das, and Amt Konar. Kerne based automatc custerng usng modfed partce swarm optmzaton agorthm. In Dr Therens and et a., edtors, Proceedngs of the Genetc and Evoutonary Computaton Conference (GECCO), pages 9, Unversty Coege London, Gower Street, London, UK, Juy 007. ACM. [] E Apaydn. Introducton to Machne Learnng (Adaptve Computaton and Machne Learnng). The MIT Press, 004. [3] P. Andras. Kerne Kohonen networs. Int. J. Neura Syst., 1: , 00. [4] Chrstopher M. Bshop. Neura Networs for Pattern Recognton. Oxford Unversty Press, [5] L. N. De Castro and J. Tmms. An artfca mmune networ for mutmoda optmsaton. In In Congress on Evoutonary Computaton. Part of the 00 IEEE Word Congress on Computatona Integence, pages , Honouu, Hawa, USA, May 00. IEEE. [6] L. N. De Castro and J. Tmms. Artfca Immune Systems: A New Computacona Integence Approach. Sprnger Verag, 00. [7] L. N. de Castro and F. J. V. Zuben. anet: An Artfca Immune Networ for Data Anayss. In H. A. Abbas R. A. S. and Chares S. Newton, edtors, Data Mnng: A Heurstc Approach, chapter XII, pages Idea Group Pubshng, USA, 001. [8] J. C. Gaeano, A. Veoza Suan, and F. A. Gonzáez. A comparatve anayss of artfca mmune networ modes. In H. Beyer, edtor, Proceedngs of GECCO 005, pages ACM Press, 005. [9] Davd E. Godberg. Genetc Agorthms n Search, Optmzaton and Machne Learnng. Addson Wesey, [10] J. E. Hunt and D. E. Cooe. Learnng usng an artfca mmune system. Journa of Networ and Computer Appcatons, 19:189 1, [11] A. Ishguro and Y. Uchawa. A gat acquston of sxegged robot usng mmune networs. In Proceedngs of the Internatona Conference on Integent Robotcs and Systems (IROS?94), voume, pages , Munch, Germany, [1] Kng Wa Lau, Huun Yn, and Smon Hubbard. Kerne seforgansng maps for cassfcaton. Neurocomputng, 69: , 005. [13] R. S. Ledey and C. Y. Suen. Pattern recognton, the ourna of the pattern recognton socety, October 006. [14] G. C. Luh and W. W. Lu. Reactve mmune networ based mobe robot navgaton. In G. NcosaV. CuteoP. J. Bentey. Tmms, edtor, Proceedng of the Thrd Conference ICARIS, pages Sprnger, 004. [15] Nea. M. Meta stabe memory n an artfca mmune networ. In P. Bentey J. Tmms and E. Hart, edtors, Proceedngs of the Second Internatona Conference ICARIS, pages , Ednburg, UK, September 003. Sprnger. [16] D. MacDonad and C. Fyfe. The erne sef organsng map. In Proc. 4th nt. conf. on nowedge based ntegence engneerng systems and apped technooges, pages , 000. [17] R. Mchean and F. J. Von Zuben. Decentrazed contro system for autonomous navgaton based on an evoved artfca mmune networ. In Proceedngs of the IEEE Congress on Evoutonary Computaton, voume, Honouu, HI, May 00. IEEE. [18] O. Nasraou, C. Cardona, C. Roas, and F. Gonzáez. Tecnostreams: Tracng evovng custers n nosy data streams wth a scaabe mmune system earnng mode. In Thrd IEEE Internatona
7 Haca un Modeo de Red Inmunoógca Artfca Basado en Kernes Gaeano, Gonzáez 71 Conference on Data Mnng,Mebourne, FL, November 003. IEEE. [19] O. Nasraou, F. Gonzáez, and D. Dasgupta. The fuzzy artfca mmune system: Motvatons, basc concepts and appcaton to custerng and web profng. In IEEE Internatona Conference on Fuzzy Systems, pages , Hawa, HI, May 00. IEEE. [0] M. Nea. An artfca mmune system for contnuous anayss of tme varyng data. In J. Tmms and P. J. Bentey, edtors, Proceedngs of the 1st Internatona Conference on Artfca Immune Systems (ICARIS), voume 1, pages 76 85, Unversty of Kent at Canterbury, September 00. Unversty of Kent at Canterbury Prntng Unt. [1] Z. S. Pan, S. C. Chen, and D. Q. Zhang. A erne base SOM cassfer n nput space. Acta Eectron. Sn., 3:7 31, 004. n Chnese. [] A. S. Pereson and G. Wesbuch. Immunoogy for physcsts. Revews of Modern Physcs, 69(4): , October [3] Bernhard Schopf and Aexander J. Smoa. Learnng wth Kernes: Support Vector Machnes, Reguarzaton, Optmzaton and Beyond. The MIT Press, December 001. [4] A. Secer, A.and Fretas and J. Tmms. Asec: An artfca mmune system for e ma cassfcaton. In H. Abbass T. Kay Chen R. McKay D. Essam R. Sarer, R. Reynods and T. Gedeon, edtors, Proceedngs of the Congress on Evoutonary Computaton, page 131? 139, Canberra. Austraa, December 003. IEEE. [5] J. Shawe Tayor and N. Crstann. Kerne Methods for Pattern Anayss. Cambrdge Unversty Press., 004. [6] J. Tmms, Nea M., and J. Hunt. An artfca mmune system for data anayss. BoSystems, 55: , 000. [7] J. Tmms and M. Nea. A resource mted artfca mmune system for data anayss. Knowedge Based Systems, 14:11 130, 001. J uan Caros Gaeano. Earned hs bacheor's degree n computer scence and engneerng from the Natona Unversty of Coomba at Bogota, and he s currenty a master student n computer scence and engneerng at the same unversty. Fabo A. Gonzáez. Earned hs bacheor's degree n computer scence and engneerng, and hs master's degree n mathematcs from the Natona Unversty of Coomba at Bogota. He aso earned a master's degree and a PhD degree n computer scence from the Unversty of Memphs, USA. He s currenty an Assocate Professor of the Systems and Computer Engneerng Department at the Nacona Unversty of Coomba.
8 7 Revsta Avances en Sstemas e Informátca, Vo.5 No. 1, Edcón Especa, Medeín, Mayo de 008, ISSN III Congreso Coombano de Computacón 3CCC 008
Multi-agent System for Custom Relationship Management with SVMs Tool
Mut-agent System for Custom Reatonshp Management wth SVMs oo Yanshan Xao, Bo Lu, 3, Dan Luo, and Longbng Cao Guangzhou Asan Games Organzng Commttee, Guangzhou 5063, P.R. Chna Facuty of Informaton echnoogy,
More informationClustering based Two-Stage Text Classification Requiring Minimal Training Data
OI: 10.2298/CSIS120130044Z Custerng based Two-Stage Text Cassfcaton Requrng Mnma Tranng ata Xue Zhang 1,2 and Wangxn Xao 3,4 1 Key Laboratory of Hgh Confdence Software Technooges, Mnstry of Educaton, Pekng
More informationAn Efficient Job Scheduling for MapReduce Clusters
Internatona Journa of Future Generaton ommuncaton and Networkng, pp. 391-398 http://dx.do.org/10.14257/jfgcn.2015.8.2.32 An Effcent Job Schedung for MapReduce usters Jun Lu 1, Tanshu Wu 1, and Mng We Ln
More informationAn Ensemble Classification Framework to Evolving Data Streams
Internatona Journa of Scence and Research (IJSR) ISSN (Onne): 39-7064 An Ensembe Cassfcaton Framework to Evovng Data Streams Naga Chthra Dev. R MCA, (M.Ph), Sr Jayendra Saraswathy Maha Vdyaaya, Coege of
More informationTCP/IP Interaction Based on Congestion Price: Stability and Optimality
TCP/IP Interacton Based on Congeston Prce: Stabty and Optmaty Jayue He Eectrca Engneerng Prnceton Unversty Ema: jhe@prncetonedu Mung Chang Eectrca Engneerng Prnceton Unversty Ema: changm@prncetonedu Jennfer
More information8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by
6 CHAPTER 8 COMPLEX VECTOR SPACES 5. Fnd the kernel of the lnear transformaton gven n Exercse 5. In Exercses 55 and 56, fnd the mage of v, for the ndcated composton, where and are gven by the followng
More informationDynamic Virtual Network Allocation for OpenFlow Based Cloud Resident Data Center
56 IEICE TRANS. COMMUN., VOL.E96 B, NO. JANUARY 203 PAPER Speca Secton on Networ Vrtuazaton, and Fuson Patform of Computng and Networng Dynamc Vrtua Networ Aocaton for OpenFow Based Coud Resdent Data Center
More informationNeural Network-based Colonoscopic Diagnosis Using On-line Learning and Differential Evolution
Neura Networ-based Coonoscopc Dagnoss Usng On-ne Learnng and Dfferenta Evouton George D. Magouas, Vasss P. Paganaos * and Mchae N. Vrahats * Department of Informaton Systems and Computng, Brune Unversty,
More informationVision Mouse. Saurabh Sarkar a* University of Cincinnati, Cincinnati, USA ABSTRACT 1. INTRODUCTION
Vson Mouse Saurabh Sarkar a* a Unversty of Cncnnat, Cncnnat, USA ABSTRACT The report dscusses a vson based approach towards trackng of eyes and fngers. The report descrbes the process of locatng the possble
More informationv a 1 b 1 i, a 2 b 2 i,..., a n b n i.
SECTION 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS 455 8.4 COMPLEX VECTOR SPACES AND INNER PRODUCTS All the vector spaces we have studed thus far n the text are real vector spaces snce the scalars are
More informationApproximation Algorithms for Data Distribution with Load Balancing of Web Servers
Approxmaton Agorthms for Data Dstrbuton wth Load Baancng of Web Servers L-Chuan Chen Networkng and Communcatons Department The MITRE Corporaton McLean, VA 22102 chen@mtreorg Hyeong-Ah Cho Department of
More informationPrediction of Success or Fail of Students on Different Educational Majors at the End of the High School with Artificial Neural Networks Methods
Predcton of Success or Fa of on Dfferent Educatona Maors at the End of the Hgh Schoo th Artfca Neura Netors Methods Sayyed Mad Maznan, Member, IACSIT, and Sayyede Azam Aboghasempur Abstract The man obectve
More informationLuby s Alg. for Maximal Independent Sets using Pairwise Independence
Lecture Notes for Randomzed Algorthms Luby s Alg. for Maxmal Independent Sets usng Parwse Independence Last Updated by Erc Vgoda on February, 006 8. Maxmal Independent Sets For a graph G = (V, E), an ndependent
More informationPredicting Advertiser Bidding Behaviors in Sponsored Search by Rationality Modeling
Predctng Advertser Bddng Behavors n Sponsored Search by Ratonaty Modeng Hafeng Xu Centre for Computatona Mathematcs n Industry and Commerce Unversty of Wateroo Wateroo, ON, Canada hafeng.ustc@gma.com Dy
More informationCardiovascular Event Risk Assessment Fusion of Individual Risk Assessment Tools Applied to the Portuguese Population
Cardovascuar Event Rsk Assessment Fuson of Indvdua Rsk Assessment Toos Apped to the Portuguese Popuaton S. Paredes, T. Rocha, P. de Carvaho, J. Henrques, J. Moras*, J. Ferrera, M. Mendes Abstract Cardovascuar
More informationThe Development of Web Log Mining Based on Improve-K-Means Clustering Analysis
The Development of Web Log Mnng Based on Improve-K-Means Clusterng Analyss TngZhong Wang * College of Informaton Technology, Luoyang Normal Unversty, Luoyang, 471022, Chna wangtngzhong2@sna.cn Abstract.
More informationOff-line and on-line scheduling on heterogeneous master-slave platforms
Laboratore de Informatque du Paraésme Écoe Normae Supéreure de Lyon Unté Mxte de Recherche CNRS-INRIA-ENS LYON-UCBL n o 5668 Off-ne and on-ne schedung on heterogeneous master-save patforms Jean-Franços
More informationSUPPORT VECTOR MACHINE FOR REGRESSION AND APPLICATIONS TO FINANCIAL FORECASTING
SUPPORT VECTOR MACHINE FOR REGRESSION AND APPICATIONS TO FINANCIA FORECASTING Theodore B. Trafas and Husen Ince Schoo of Industra Engneerng Unverst of Okahoma W. Bod Sute 4 Norman Okahoma 739 trafas@ecn.ou.edu;
More informationExpressive Negotiation over Donations to Charities
Expressve Negotaton over Donatons to Chartes Vncent Contzer Carnege Meon Unversty 5000 Forbes Avenue Pttsburgh, PA 523, USA contzer@cs.cmu.edu Tuomas Sandhom Carnege Meon Unversty 5000 Forbes Avenue Pttsburgh,
More informationGRADIENT METHODS FOR BINARY INTEGER PROGRAMMING
Proceedns of the 4st Internatona Conference on Computers & Industra Enneern GRADIENT METHODS FOR BINARY INTEGER PROGRAMMING Chen-Yuan Huan¹, Ta-Chun Wan² Insttute of Cv Avaton, Natona Chen Kun Unversty,
More informationImplementation of Deutsch's Algorithm Using Mathcad
Implementaton of Deutsch's Algorthm Usng Mathcad Frank Roux The followng s a Mathcad mplementaton of Davd Deutsch's quantum computer prototype as presented on pages - n "Machnes, Logc and Quantum Physcs"
More information"Research Note" APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES *
Iranan Journal of Scence & Technology, Transacton B, Engneerng, ol. 30, No. B6, 789-794 rnted n The Islamc Republc of Iran, 006 Shraz Unversty "Research Note" ALICATION OF CHARGE SIMULATION METHOD TO ELECTRIC
More informationA Simple Congestion-Aware Algorithm for Load Balancing in Datacenter Networks
A Smpe Congeston-Aware Agorthm for Load Baancng n Datacenter Networs Mehrnoosh Shafee, and Javad Ghader, Coumba Unversty Abstract We study the probem of oad baancng n datacenter networs, namey, assgnng
More informationA Resources Allocation Model for Multi-Project Management
A Resources Aocaton Mode for Mut-Proect Management Hamdatou Kane, Aban Tsser To cte ths verson: Hamdatou Kane, Aban Tsser. A Resources Aocaton Mode for Mut-Proect Management. 9th Internatona Conference
More informationPredictive Control of a Smart Grid: A Distributed Optimization Algorithm with Centralized Performance Properties*
Predctve Contro of a Smart Grd: A Dstrbuted Optmzaton Agorthm wth Centrazed Performance Propertes* Phpp Braun, Lars Grüne, Chrstopher M. Keett 2, Steven R. Weer 2, and Kar Worthmann 3 Abstract The authors
More informationWhat is Candidate Sampling
What s Canddate Samplng Say we have a multclass or mult label problem where each tranng example ( x, T ) conssts of a context x a small (mult)set of target classes T out of a large unverse L of possble
More informationSupport Vector Machines
Support Vector Machnes Max Wellng Department of Computer Scence Unversty of Toronto 10 Kng s College Road Toronto, M5S 3G5 Canada wellng@cs.toronto.edu Abstract Ths s a note to explan support vector machnes.
More informationA Simple Approach to Clustering in Excel
A Smple Approach to Clusterng n Excel Aravnd H Center for Computatonal Engneerng and Networng Amrta Vshwa Vdyapeetham, Combatore, Inda C Rajgopal Center for Computatonal Engneerng and Networng Amrta Vshwa
More informationAn Interest-Oriented Network Evolution Mechanism for Online Communities
An Interest-Orented Network Evoluton Mechansm for Onlne Communtes Cahong Sun and Xaopng Yang School of Informaton, Renmn Unversty of Chna, Bejng 100872, P.R. Chna {chsun,yang}@ruc.edu.cn Abstract. Onlne
More informationSIMPLIFYING NDA PROGRAMMING WITH PROt SQL
SIMPLIFYING NDA PROGRAMMING WITH PROt SQL Aeen L. Yam, Besseaar Assocates, Prnceton, NJ ABSRACf The programmng of New Drug Appcaton (NDA) Integrated Summary of Safety (ISS) usuay nvoves obtanng patent
More informationProject Networks With Mixed-Time Constraints
Project Networs Wth Mxed-Tme Constrants L Caccetta and B Wattananon Western Australan Centre of Excellence n Industral Optmsaton (WACEIO) Curtn Unversty of Technology GPO Box U1987 Perth Western Australa
More informationFactored Conditional Restricted Boltzmann Machines for Modeling Motion Style
Factored Condtona Restrcted Botzmann Machnes for Modeng Moton Stye Graham W. Tayor GWTAYLOR@CS.TORONTO.EDU Geoffrey E. Hnton HINTON@CS.TORONTO.EDU Department of Computer Scence, Unversty of Toronto, Toronto,
More informationExtending Probabilistic Dynamic Epistemic Logic
Extendng Probablstc Dynamc Epstemc Logc Joshua Sack May 29, 2008 Probablty Space Defnton A probablty space s a tuple (S, A, µ), where 1 S s a set called the sample space. 2 A P(S) s a σ-algebra: a set
More informationFace Verification Problem. Face Recognition Problem. Application: Access Control. Biometric Authentication. Face Verification (1:1 matching)
Face Recognton Problem Face Verfcaton Problem Face Verfcaton (1:1 matchng) Querymage face query Face Recognton (1:N matchng) database Applcaton: Access Control www.vsage.com www.vsoncs.com Bometrc Authentcaton
More informationFeature selection for intrusion detection. Slobodan Petrović NISlab, Gjøvik University College
Feature selecton for ntruson detecton Slobodan Petrovć NISlab, Gjøvk Unversty College Contents The feature selecton problem Intruson detecton Traffc features relevant for IDS The CFS measure The mrmr measure
More informationBERNSTEIN POLYNOMIALS
On-Lne Geometrc Modelng Notes BERNSTEIN POLYNOMIALS Kenneth I. Joy Vsualzaton and Graphcs Research Group Department of Computer Scence Unversty of Calforna, Davs Overvew Polynomals are ncredbly useful
More informationAn Enhanced Super-Resolution System with Improved Image Registration, Automatic Image Selection, and Image Enhancement
An Enhanced Super-Resoluton System wth Improved Image Regstraton, Automatc Image Selecton, and Image Enhancement Yu-Chuan Kuo ( ), Chen-Yu Chen ( ), and Chou-Shann Fuh ( ) Department of Computer Scence
More informationSupport vector domain description
Pattern Recognton Letters 20 (1999) 1191±1199 www.elsever.nl/locate/patrec Support vector doman descrpton Davd M.J. Tax *,1, Robert P.W. Dun Pattern Recognton Group, Faculty of Appled Scence, Delft Unversty
More information2) A single-language trained classifier: one. classifier trained on documents written in
Openng the ega terature Porta to mutngua access E. Francescon, G. Perugne ITTIG Insttute of Lega Informaton Theory and Technooges Itaan Natona Research Counc, Forence, Itay Te: +39 055 43999 Fax: +39 055
More informationOn-Line Trajectory Generation: Nonconstant Motion Constraints
2012 IEEE Internatona Conference on Robotcs and Automaton RverCentre, Sant Pau, Mnnesota, USA May 14-18, 2012 On-Lne Trajectory Generaton: Nonconstant Moton Constrants Torsten Kröger Abstract A concept
More informationAsymptotically Optimal Inventory Control for Assemble-to-Order Systems with Identical Lead Times
Asymptotcay Optma Inventory Contro for Assembe-to-Order Systems wth Identca ead Tmes Martn I. Reman Acate-ucent Be abs, Murray H, NJ 07974, marty@research.be-abs.com Qong Wang Industra and Enterprse Systems
More informationA Statistical Model for Detecting Abnormality in Static-Priority Scheduling Networks with Differentiated Services
A Statstcal odel for Detectng Abnoralty n Statc-Prorty Schedulng Networks wth Dfferentated Servces ng L 1 and We Zhao 1 School of Inforaton Scence & Technology, East Chna Noral Unversty, Shangha 0006,
More informationForecasting the Demand of Emergency Supplies: Based on the CBR Theory and BP Neural Network
700 Proceedngs of the 8th Internatonal Conference on Innovaton & Management Forecastng the Demand of Emergency Supples: Based on the CBR Theory and BP Neural Network Fu Deqang, Lu Yun, L Changbng School
More informationFault tolerance in cloud technologies presented as a service
Internatonal Scentfc Conference Computer Scence 2015 Pavel Dzhunev, PhD student Fault tolerance n cloud technologes presented as a servce INTRODUCTION Improvements n technques for vrtualzaton and performance
More informationModule 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur
Module LOSSLESS IMAGE COMPRESSION SYSTEMS Lesson 3 Lossless Compresson: Huffman Codng Instructonal Objectves At the end of ths lesson, the students should be able to:. Defne and measure source entropy..
More informationAn Analysis of Dynamic Severity and Population Size
An Analyss of Dynamc Severty and Populaton Sze Karsten Wecker Unversty of Stuttgart, Insttute of Computer Scence, Bretwesenstr. 2 22, 7565 Stuttgart, Germany, emal: Karsten.Wecker@nformatk.un-stuttgart.de
More informationRecurrence. 1 Definitions and main statements
Recurrence 1 Defntons and man statements Let X n, n = 0, 1, 2,... be a MC wth the state space S = (1, 2,...), transton probabltes p j = P {X n+1 = j X n = }, and the transton matrx P = (p j ),j S def.
More informationA machine vision approach for detecting and inspecting circular parts
A machne vson approach for detectng and nspectng crcular parts Du-Mng Tsa Machne Vson Lab. Department of Industral Engneerng and Management Yuan-Ze Unversty, Chung-L, Tawan, R.O.C. E-mal: edmtsa@saturn.yzu.edu.tw
More informationInstitute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic
Lagrange Multplers as Quanttatve Indcators n Economcs Ivan Mezník Insttute of Informatcs, Faculty of Busness and Management, Brno Unversty of TechnologCzech Republc Abstract The quanttatve role of Lagrange
More informationA Secure Password-Authenticated Key Agreement Using Smart Cards
A Secure Password-Authentcated Key Agreement Usng Smart Cards Ka Chan 1, Wen-Chung Kuo 2 and Jn-Chou Cheng 3 1 Department of Computer and Informaton Scence, R.O.C. Mltary Academy, Kaohsung 83059, Tawan,
More informationLogistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification
Lecture 4: More classfers and classes C4B Machne Learnng Hlary 20 A. Zsserman Logstc regresson Loss functons revsted Adaboost Loss functons revsted Optmzaton Multple class classfcaton Logstc Regresson
More informationResearch on Single and Mixed Fleet Strategy for Open Vehicle Routing Problem
276 JOURNAL OF SOFTWARE, VOL 6, NO, OCTOBER 2 Research on Snge and Mxed Feet Strategy for Open Vehce Routng Probe Chunyu Ren Heongjang Unversty /Schoo of Inforaton scence and technoogy, Harbn, Chna Ea:
More informationAnts Can Schedule Software Projects
Ants Can Schedule Software Proects Broderck Crawford 1,2, Rcardo Soto 1,3, Frankln Johnson 4, and Erc Monfroy 5 1 Pontfca Unversdad Católca de Valparaíso, Chle FrstName.Name@ucv.cl 2 Unversdad Fns Terrae,
More informationA Fast Incremental Spectral Clustering for Large Data Sets
2011 12th Internatonal Conference on Parallel and Dstrbuted Computng, Applcatons and Technologes A Fast Incremental Spectral Clusterng for Large Data Sets Tengteng Kong 1,YeTan 1, Hong Shen 1,2 1 School
More informationSwing-Free Transporting of Two-Dimensional Overhead Crane Using Sliding Mode Fuzzy Control
Swng-Free Transportng of Two-Dmensona Overhead Crane Usng Sdng Mode Fuzzy Contro Dantong Lu, Janqang, Dongn Zhao, and We Wang Astract An adaptve sdng mode fuzzy contro approach s proposed for a two-dmensona
More informationThe OC Curve of Attribute Acceptance Plans
The OC Curve of Attrbute Acceptance Plans The Operatng Characterstc (OC) curve descrbes the probablty of acceptng a lot as a functon of the lot s qualty. Fgure 1 shows a typcal OC Curve. 10 8 6 4 1 3 4
More informationAdaptive Multi-Compositionality for Recursive Neural Models with Applications to Sentiment Analysis
Proceedngs of the Twenty-Eghth AAAI Conference on Artfca Integence Adapte Mut-Compostonaty for Recurse Neura Modes wth Appcatons to Sentment Anayss L Dong Furu We Mng Zhou Ke Xu State Key Lab of Software
More informationLoop Parallelization
- - Loop Parallelzaton C-52 Complaton steps: nested loops operatng on arrays, sequentell executon of teraton space DECLARE B[..,..+] FOR I :=.. FOR J :=.. I B[I,J] := B[I-,J]+B[I-,J-] ED FOR ED FOR analyze
More informationGender Classification for Real-Time Audience Analysis System
Gender Classfcaton for Real-Tme Audence Analyss System Vladmr Khryashchev, Lev Shmaglt, Andrey Shemyakov, Anton Lebedev Yaroslavl State Unversty Yaroslavl, Russa vhr@yandex.ru, shmaglt_lev@yahoo.com, andrey.shemakov@gmal.com,
More informationDescriptive Models. Cluster Analysis. Example. General Applications of Clustering. Examples of Clustering Applications
CMSC828G Prncples of Data Mnng Lecture #9 Today s Readng: HMS, chapter 9 Today s Lecture: Descrptve Modelng Clusterng Algorthms Descrptve Models model presents the man features of the data, a global summary
More informationMACHINE VISION SYSTEM FOR SPECULAR SURFACE INSPECTION: USE OF SIMULATION PROCESS AS A TOOL FOR DESIGN AND OPTIMIZATION
MACHINE VISION SYSTEM FOR SPECULAR SURFACE INSPECTION: USE OF SIMULATION PROCESS AS A TOOL FOR DESIGN AND OPTIMIZATION R. SEULIN, F. MERIENNE and P. GORRIA Laboratore Le2, CNRS FRE2309, EA 242, Unversté
More informationBranch-and-Price and Heuristic Column Generation for the Generalized Truck-and-Trailer Routing Problem
REVISTA DE MÉTODOS CUANTITATIVOS PARA LA ECONOMÍA Y LA EMPRESA (12) Págnas 5 38 Dcembre de 2011 ISSN: 1886-516X DL: SE-2927-06 URL: http://wwwupoes/revmetcuant/artphp?d=51 Branch-and-Prce and Heurstc Coumn
More informationSection 5.3 Annuities, Future Value, and Sinking Funds
Secton 5.3 Annutes, Future Value, and Snkng Funds Ordnary Annutes A sequence of equal payments made at equal perods of tme s called an annuty. The tme between payments s the payment perod, and the tme
More informationRing structure of splines on triangulations
www.oeaw.ac.at Rng structure of splnes on trangulatons N. Vllamzar RICAM-Report 2014-48 www.rcam.oeaw.ac.at RING STRUCTURE OF SPLINES ON TRIANGULATIONS NELLY VILLAMIZAR Introducton For a trangulated regon
More informationA Multi-Camera System on PC-Cluster for Real-time 3-D Tracking
The 23 rd Conference of the Mechancal Engneerng Network of Thaland November 4 7, 2009, Chang Ma A Mult-Camera System on PC-Cluster for Real-tme 3-D Trackng Vboon Sangveraphunsr*, Krtsana Uttamang, and
More informationThe Dynamics of Wealth and Income Distribution in a Neoclassical Growth Model * Stephen J. Turnovsky. University of Washington, Seattle
The Dynamcs of Weath and Income Dstrbuton n a Neocassca Growth Mode * Stephen J. Turnovsy Unversty of Washngton, Seatte Ceca García-Peñaosa CNRS and GREQAM March 26 Abstract: We examne the evouton of the
More informationDistributed Multi-Target Tracking In A Self-Configuring Camera Network
Dstrbuted Mult-Target Trackng In A Self-Confgurng Camera Network Crstan Soto, B Song, Amt K. Roy-Chowdhury Department of Electrcal Engneerng Unversty of Calforna, Rversde {cwlder,bsong,amtrc}@ee.ucr.edu
More informationwhere the coordinates are related to those in the old frame as follows.
Chapter 2 - Cartesan Vectors and Tensors: Ther Algebra Defnton of a vector Examples of vectors Scalar multplcaton Addton of vectors coplanar vectors Unt vectors A bass of non-coplanar vectors Scalar product
More informationReview of Hierarchical Models for Data Clustering and Visualization
Revew of Herarchcal Models for Data Clusterng and Vsualzaton Lola Vcente & Alfredo Velldo Grup de Soft Computng Seccó d Intel lgènca Artfcal Departament de Llenguatges Sstemes Informàtcs Unverstat Poltècnca
More information7.5. Present Value of an Annuity. Investigate
7.5 Present Value of an Annuty Owen and Anna are approachng retrement and are puttng ther fnances n order. They have worked hard and nvested ther earnngs so that they now have a large amount of money on
More informationDesign of Output Codes for Fast Covering Learning using Basic Decomposition Techniques
Journal of Computer Scence (7): 565-57, 6 ISSN 59-66 6 Scence Publcatons Desgn of Output Codes for Fast Coverng Learnng usng Basc Decomposton Technques Aruna Twar and Narendra S. Chaudhar, Faculty of Computer
More informationCalculation of Sampling Weights
Perre Foy Statstcs Canada 4 Calculaton of Samplng Weghts 4.1 OVERVIEW The basc sample desgn used n TIMSS Populatons 1 and 2 was a two-stage stratfed cluster desgn. 1 The frst stage conssted of a sample
More informationUSING EMPIRICAL LIKELIHOOD TO COMBINE DATA: APPLICATION TO FOOD RISK ASSESSMENT.
Submtted to the Annas of Apped Statstcs USING EMPIRICA IKEIHOOD TO COMBINE DATA: APPICATION TO FOOD RISK ASSESSMENT. By Amée Crépet, Hugo Harar-Kermadec and Jessca Tressou INRA Mét@rs and INRA COREA Ths
More informationTraffic State Estimation in the Traffic Management Center of Berlin
Traffc State Estmaton n the Traffc Management Center of Berln Authors: Peter Vortsch, PTV AG, Stumpfstrasse, D-763 Karlsruhe, Germany phone ++49/72/965/35, emal peter.vortsch@ptv.de Peter Möhl, PTV AG,
More informationTHE APPLICATION OF DATA MINING TECHNIQUES AND MULTIPLE CLASSIFIERS TO MARKETING DECISION
Internatonal Journal of Electronc Busness Management, Vol. 3, No. 4, pp. 30-30 (2005) 30 THE APPLICATION OF DATA MINING TECHNIQUES AND MULTIPLE CLASSIFIERS TO MARKETING DECISION Yu-Mn Chang *, Yu-Cheh
More informationThe Greedy Method. Introduction. 0/1 Knapsack Problem
The Greedy Method Introducton We have completed data structures. We now are gong to look at algorthm desgn methods. Often we are lookng at optmzaton problems whose performance s exponental. For an optmzaton
More informationAn interactive system for structure-based ASCII art creation
An nteractve system for structure-based ASCII art creaton Katsunor Myake Henry Johan Tomoyuk Nshta The Unversty of Tokyo Nanyang Technologcal Unversty Abstract Non-Photorealstc Renderng (NPR), whose am
More informationXAC08-6 Professional Project Management
1 XAC08-6 Professona Project anagement Ths Lecture: Tte s so manager shoud ncude a s management pan a document that gudes any experts agree Some faure project to Ba, ba, ba, ba Communcaton anagement Wee
More informationPSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 12
14 The Ch-squared dstrbuton PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 1 If a normal varable X, havng mean µ and varance σ, s standardsed, the new varable Z has a mean 0 and varance 1. When ths standardsed
More informationA Binary Particle Swarm Optimization Algorithm for Lot Sizing Problem
Journal o Economc and Socal Research 5 (2), -2 A Bnary Partcle Swarm Optmzaton Algorthm or Lot Szng Problem M. Fath Taşgetren & Yun-Cha Lang Abstract. Ths paper presents a bnary partcle swarm optmzaton
More informationHowHow to Find the Best Online Stock Broker
A GENERAL APPROACH FOR SECURITY MONITORING AND PREVENTIVE CONTROL OF NETWORKS WITH LARGE WIND POWER PRODUCTION Helena Vasconcelos INESC Porto hvasconcelos@nescportopt J N Fdalgo INESC Porto and FEUP jfdalgo@nescportopt
More informationHow To Create An Emoton Recognzer
Recognzng Low/Hgh Anger n Speech for Call Centers FU-MING LEE*, LI-HUA LI, RU-YI HUANG Department of Informaton Management Chaoyang Unversty of Technology 168 Jfong E. Rd., Wufong Townshp Tachung County,
More informationbenefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ).
REVIEW OF RISK MANAGEMENT CONCEPTS LOSS DISTRIBUTIONS AND INSURANCE Loss and nsurance: When someone s subject to the rsk of ncurrng a fnancal loss, the loss s generally modeled usng a random varable or
More informationSimple Interest Loans (Section 5.1) :
Chapter 5 Fnance The frst part of ths revew wll explan the dfferent nterest and nvestment equatons you learned n secton 5.1 through 5.4 of your textbook and go through several examples. The second part
More informationGRAVITY DATA VALIDATION AND OUTLIER DETECTION USING L 1 -NORM
GRAVITY DATA VALIDATION AND OUTLIER DETECTION USING L 1 -NORM BARRIOT Jean-Perre, SARRAILH Mchel BGI/CNES 18.av.E.Beln 31401 TOULOUSE Cedex 4 (France) Emal: jean-perre.barrot@cnes.fr 1/Introducton The
More informationCausal, Explanatory Forecasting. Analysis. Regression Analysis. Simple Linear Regression. Which is Independent? Forecasting
Causal, Explanatory Forecastng Assumes cause-and-effect relatonshp between system nputs and ts output Forecastng wth Regresson Analyss Rchard S. Barr Inputs System Cause + Effect Relatonshp The job of
More informationECE544NA Final Project: Robust Machine Learning Hardware via Classifier Ensemble
1 ECE544NA Fnal Project: Robust Machne Learnng Hardware va Classfer Ensemble Sa Zhang, szhang12@llnos.edu Dept. of Electr. & Comput. Eng., Unv. of Illnos at Urbana-Champagn, Urbana, IL, USA Abstract In
More informationCFD Simulation of Cloud and Tip Vortex Cavitation on Hydrofoils
ICMF 007, Lepzg, Germany, Juy 9 3, 007 CFD Smuaton of Coud and Tp Vortex Cavtaton on Hydrofos Th. Fran, C. Lfante, S. Jebauer, M. Kuntz and K. Rec ) ANSYS Germany, CFX Deveopment Staudenfedweg, D-8364
More informationImplementations of Web-based Recommender Systems Using Hybrid Methods
Internatonal Journal of Computer Scence & Applcatons Vol. 3 Issue 3, pp 52-64 2006 Technomathematcs Research Foundaton Implementatons of Web-based Recommender Systems Usng Hybrd Methods Janusz Sobeck Insttute
More informationConversion between the vector and raster data structures using Fuzzy Geographical Entities
Converson between the vector and raster data structures usng Fuzzy Geographcal Enttes Cdála Fonte Department of Mathematcs Faculty of Scences and Technology Unversty of Combra, Apartado 38, 3 454 Combra,
More informationResearch on Transformation Engineering BOM into Manufacturing BOM Based on BOP
Appled Mechancs and Materals Vols 10-12 (2008) pp 99-103 Onlne avalable snce 2007/Dec/06 at wwwscentfcnet (2008) Trans Tech Publcatons, Swtzerland do:104028/wwwscentfcnet/amm10-1299 Research on Transformaton
More informationA Decision Support System for Crowd Control
A Decson Support System for Crowd Control Johan Schubert, Lug Ferrara, Pontus Hörlng, Johan Walter Department of Decson Support Systems Dvson of Command and Control Systems Swedsh Defence Research Agency
More informationIncreasing Supported VoIP Flows in WMNs through Link-Based Aggregation
Increasng Supported VoIP Fows n WMNs through n-based Aggregaton J. Oech, Y. Hamam, A. Kuren F SATIE TUT Pretora, South Afrca oechr@gma.com T. Owa Meraa Insttute Counc of Scentfc and Industra Research (CSIR)
More information) of the Cell class is created containing information about events associated with the cell. Events are added to the Cell instance
Calbraton Method Instances of the Cell class (one nstance for each FMS cell) contan ADC raw data and methods assocated wth each partcular FMS cell. The calbraton method ncludes event selecton (Class Cell
More informationTexas Instruments 30X IIS Calculator
Texas Instruments 30X IIS Calculator Keystrokes for the TI-30X IIS are shown for a few topcs n whch keystrokes are unque. Start by readng the Quk Start secton. Then, before begnnng a specfc unt of the
More informationSoftware project management with GAs
Informaton Scences 177 (27) 238 241 www.elsever.com/locate/ns Software project management wth GAs Enrque Alba *, J. Francsco Chcano Unversty of Málaga, Grupo GISUM, Departamento de Lenguajes y Cencas de
More informationTesting CAB-IDS through Mutations: on the Identification of Network Scans
Testng CAB-IDS through Mutatons: on the Identfcaton of Network Scans Emlo Corchado, Álvaro Herrero, José Manuel Sáz Department of Cvl Engneerng, Unversty of Burgos, Span {escorchado, ahcoso, msaz}@ubu.es
More informationUsing an Adaptive Fuzzy Logic System to Optimise Knowledge Discovery in Proteomics
Usng an Adaptve Fuzzy Logc System to Optmse Knowledge Dscovery n Proteomcs James Malone, Ken McGarry and Chrs Bowerman School of Computng and Technology Sunderland Unversty St. Peter s Way, Sunderland,
More informationMining Feature Importance: Applying Evolutionary Algorithms within a Web-based Educational System
Mnng Feature Importance: Applyng Evolutonary Algorthms wthn a Web-based Educatonal System Behrouz MINAEI-BIDGOLI 1, and Gerd KORTEMEYER 2, and Wllam F. PUNCH 1 1 Genetc Algorthms Research and Applcatons
More information