Joint Probabilistic Curve Clustering and Alignment

Size: px
Start display at page:

Download "Joint Probabilistic Curve Clustering and Alignment"

Transcription

1 Jont Probablstc Curve Clusterng and Algnment Scott Gaffney and Padhrac Smyth School of Informaton and Computer Scence Unversty of Calforna, Irvne, CA Abstract Clusterng and predcton of sets of curves s an mportant problem n many areas of scence and engneerng. It s often the case that curves tend to be msalgned from each other n a contnuous manner, ether n space (across the measurements) or n tme. We develop a probablstc framework that allows for the jont clusterng and contnuous algnment of sets of curves n curve space (as opposed to a fxed-dmensonal feature-vector space). The proposed methodology ntegrates new probablstc algnment models wth model-based curve clusterng algorthms. The probablstc approach allows for the dervaton of consstent EMtype learnng algorthms for the jont clusterng-algnment problem. Expermental results are shown for algnment of human growth data and jont clusterng and algnment of gene expresson tme-course data. Introducton We ntroduce a novel methodology for the clusterng and predcton of sets of smoothly varyng curves whle jontly allowng for the learnng of sets of contnuous curve transformatons. Our approach s to formulate models for both the clusterng and algnment sub-problems and ntegrate them nto a unfed probablstc framework that allows for the dervaton of consstent learnng algorthms. The algnment sub-problem s handled wth the ntroducton of a novel curve algnment procedure employng model prors over the set of possble algnments leadng to the dervaton of EM learnng algorthms that formalze the so-called Procrustes approach for curve data []. These algnment models are then ntegrated nto a fnte mxture model settng n whch the clusterng s carred out. We make use of both polynomal and splne regresson mxture models to complete the jont clusterng-algnment framework. The followng smple llustratve example demonstrates the mportance of jontly handlng the clusterng-algnment problem as opposed to treatng algnment and clusterng separately. Fgure (a) shows a smulated set of curves whch have been subjected to random translatons n tme. The underlyng generatve model contans three clusters each descrbed by a cubc polynomal (not shown). Fgure (b) shows the output of the EM algorthm descrbed later n ths paper, where curves have been smultaneously algned and clustered. Ths fgure s vrtually dentcal

2 Y axs Tme Y axs Tme (a) Smulated data (b) Jont EM results Y axs Tme Y axs Tme (c) Cluster frst (d) Algn second Fgure : Comparson of jont EM and sequental clusterng-algnment: (top-row) unlabelled smulated data wth hdden algnments, and soluton recovered by EM, (bottom-row) cluster and then algn. to that of the orgnal data (wth cluster labels and no msalgnnment). Fgures (c) shows the result of frst clusterng the unalgned data, and Fgure (d) then shows the result of algnng wthn each cluster. The sequental approach results n sgnfcant msclassfcaton and ncorrect algnment, demonstratng that a two-stage approach can be qute suboptmal when compared to a jont clusterng-algnment methodology. (Smlar results, not shown, are obtaned when the curves are frst algned and then clustered see [] for full detals.) There has been lttle pror work on the specfc problem of jont curve clusterng and algnment, but there s related work n other areas. For example, clusterng of gene-expresson tme profles wth mxtures of splnes s addressed n [3]. However, algnment s only consdered as a post-processng step to compare cluster results among related datasets. In mage analyss, the transformed mxture of Gaussans (TMG) model uses a probablstc framework and an EM algorthm to jontly learn clusterng and algnment of mage patches subject to varous forms of lnear transformatons [4]. However, ths model only consders sets of transformatons n dscrete pxel space, whereas we are focused on curve modellng that allows for arbtrary contnuous algnment n tme and space. Another branch of work n mage analyss focuses on the problem of estmatng correspondences of ponts across mages [5] (or vertces across graphs [6]), usng EM or determnstc annealng algorthms. The results we descrbe here dffer prmarly n that (a) we focus specfcally on sets of curves rather than mage data (generally makng the problem more tractable), (b) we focus on clusterng and algnment rather than just algnment, (c) we allow contnuous affne transformatons n tme and measurement space, and (d) we have a fully generatve probablstc framework allowng for (for example) the ncorporaton of nformatve prors on transformatons f such pror nformaton exsts. In earler related work we developed general technques for curve clusterng (e.g., [7]) and also proposed technques for transformaton-nvarant curve clusterng wth dscrete tme algnment and Gaussan mxture models for curves [8, 9]. In ths paper we provde

3 a much more general framework that allows for contnuous algnment n both tme and measurement space for a general class of cluster shape models, ncludng polynomals and splnes. Jont clusterng and algnment It s useful to represent curves as varable-length vectors. In ths case, y s a curve that conssts of a sequence of n observatons or measurements. The j-th measurement of y s denoted by y j and s usually taken to be unvarate (the generalzaton to multvarate observatons s straghtforward). The assocated covarate of y s wrtten as x n the same manner. x s often thought of as tme so that x j gves the tme at whch y j was observed. Regresson mxture models can be effectvely used to cluster ths type of curve data []. In the standard setup, y s modelled usng a normal (Gaussan) regresson model n whch y = X β+ɛ, where β s a (p+) coeffcent vector, ɛ s a zero-mean Gaussan nose varable, and X s the regresson matrx. The form of X depends on the type of regresson model employed. For polynomal regresson, X s often assocated wth the standard Vandermonde matrx; and for splne regresson, X takes the form of a splne-bass matrx (see, e.g., [7] for more detals). The mxture model s completed by repeatng ths model over K clusters and ndexng the parameters by k so that, for example, y = X β k + ɛ gves the regresson model for y under the k-th cluster. B-splnes [] are partcularly effcent for computatonal purposes due to the blockdagonal bass matrces that result. Usng B-splnes, the curve pont y j can be represented as the lnear combnaton y j = B j c, n whch the vector B j gves the vector of B-splne bass functons evaluated at x j, and c gves the splne coeffcent vector []. The full curve y can then be wrtten compactly as y = B c n whch the splne bass matrx takes the form B =[B B n ]. Splne regresson models can be easly ntegrated nto the regresson mxture model framework by equatng the regresson matrx X wth the splne bass matrx B. In what follows, we use the more general notaton X n favor of the more specfc B.. Jont model defnton The jont clusterng-algnment model defnton s based on a regresson mxture model that has been augmented wth up to four ndvdual random transformaton parameters or varables (a,b,c,d ). The a and b allow for scalng and translaton n tme, whle the c and d allow for scalng and translaton n measurement space. The model defnton takes the form y = c a x b β k + d + ɛ, () n whch a x b represents the regresson matrx X (ether splne or polynomal) evaluated at the transformed tme a x b. Below we use the matrx X to denote a x b when parsmony s requred. It s assumed that ɛ s a zero-mean Gaussan vector wth covarance σk I. The condtonal densty p k (y a,b,c,d )=N(y c a x b β k + d,σki) () gves the probablty densty of y when all the transformaton parameters (as well as cluster membershp) are known. (Note that the densty on the left s mplctly condtoned on an approprate set of parameters ths s always assumed n what follows.) In general, the values for the transformaton parameters are unknown. Treatng ths as a standard hdden-data problem, t s useful to thnk of each of the transformaton parameters as random varables that are curve-specfc but wth populaton-level pror probablty dstrbutons. In ths

4 way, the transformaton parameters and the model parameters can be learned smultaneously n an effcent manner usng EM.. Transformaton prors Prors are attached to each of the transformaton varables n such a way that the dentty transformaton s the most lkely transformaton. A useful pror for ths s the Gaussan densty N (µ, σ ) wth mean µ and varance σ. The tme transformaton prors are specfed as a N(,r k), b N(,s k), (3) and the measurement space prors are gven as c N(,u k),d N(,v k). (4) Note that the dentty transformaton s ndeed the most lkely. All of the varance parameters are cluster-specfc n general; however, any subset of these parameters can be ted across clusters f desred n a specfc applcaton. Note that these prors techncally allow for negatve scalng n tme and n measurement space. In practce ths s typcally not a problem, though one can easly specfy other prors (e.g., log-normal) to strctly dsallow ths possblty. It should be noted that each of the pror varance parameters are learned from the data n the ensung EM algorthm. However, below we do not make use of hyperprors for these pror parameters (ths can be ntegrated n a straghtforward manner f desred)..3 Full probablty model The jont densty of y and the set of transformaton varables Φ = {a,b,c,d } can be wrtten succnctly as p k (y, Φ )=p k (y Φ )p k (Φ ), (5) where p k (Φ )=N(a,rk )N (b,s k )N (c,u k )N (d,vk ). The space transformaton parameters can be ntegrated-out of (5) resultng n the margnal of y condtoned only on the tme transformaton parameters. Ths condtonal margnal takes the form p k (y a,b ) = p k (y,c,d a,b ) dc,dd = N (y X β k, U k + V k σki), (6) wth U k = u k X β k β k X + σ k I and V k = vk + σ ki. The uncondtonal (though, stll cluster-dependent) margnal for y cannot be computed analytcally snce a,b cannot be analytcally ntegrated-out. Instead, we use numercal Monte Carlo ntegraton for ths task. The resultng uncondtonal margnal for y can be approxmated by p k (y ) = p k (y a,b )p k (a )p k (b ) da db p k (y M a (m),b (m) ), m (7) where the M Monte Carlo samples are taken accordng to a (m) N(,rk ), and b(m) N(,s k ), for m =,...,M. (8) A mxture results when cluster membershp s unknown: p(y )= α k p k (y ). k (9) The log-lkelhood of all n curves Y = {y } follows drectly from ths approxmaton and takes the form log p(y ) log mk α k p k (y a (m),b (m) ) n log M. ()

5 .4 EM algorthm We derve an EM algorthm that smultaneously allows the learnng of both the model parameters and the transformaton varables Φ wth tme-complexty that s lnear n the total number of data ponts N = n. Frst, let z gve the cluster membershp for curve y. Now, regard the transformaton varables {Φ } as well as the cluster membershps {z } as beng hdden. The complete-data log-lkelhood functon s defned as the jont loglkelhood of Y and the hdden data {Φ,z }. Ths can be wrtten as the sum over all n curves of the log of the product of α z and the cluster-dependent jont densty n (5). Ths functon takes the form L c = log α z p z (y Φ ) p z (Φ ). () In the E-step, the posteror p(φ,z y ) s calculated and then used to take the posteror expectaton of Equaton (). Ths expectaton s then used n the M-step to calculate the re-estmaton equatons for updatng the model parameters {β k,σ k,r k,s k,u k,v k }..5 E-step The posteror p(φ,z y ) can be factorzed as p z (Φ y )p(z y ). The second factor s the membershp probablty w k that y was generated by cluster k. It can be rewrtten as p(z = k y ) p k (y ) and evaluated usng Equaton (7). The frst factor requres a bt more work. Further factorng reveals that p z (Φ y )=p z (c,d a,b, y )p z (a,b y ). The new frst factor p z (c,d a,b, y ) can be solved for exactly by notng that t s proportonal to a bvarate normal dstrbuton for each z []. The new second factor p z (a,b y ) cannot, n general, be solved for analytcally, so nstead we use an approxmaton. The fact that posteror denstes tend towards hghly peaked Gaussan denstes has been wdely noted (e.g, []) and leads to the normal approxmaton of posteror denstes. To make the approxmaton here, the vector (â k, ˆb k ) representng the mult-dmensonal mode of p k (a,b y ), the covarance matrx V (k) a b for (â k, ˆb k ), and the separate varances V ak,v bk must be found. These can readly be estmated usng a Nelder-Mead optmzaton method. Experments have shown ths approxmaton works well across a varety of expermental and real-world data sets []. The above calculatons of the posteror p(φ,z y ) allow the posteror expectaton of the complete-data log-lkelhood n Equaton () to be solved for. Ths expectaton results n the so-called Q-functon whch s maxmzed n the M-step. Although the dervaton s qute complex, the Q-functon can be calculated exactly for polynomal regresson []; for splne regresson, the bass functons do not afford an exact formula for the soluton of the Q-functon. However, n the splne case, removal of a few problematc varance terms gves an effcent approxmaton (the nterested reader s referred to [] for more detals)..6 M-step The M-step s straghtforward snce most of the hard work s done n the E-step. The Q- functon s maxmzed over the set of parameters {β k,σk,r k,s k,u k,v k } for k K. The derved solutons are as follows: ˆr k = [â ] w w k k + V ak, ŝ ] k = k w w k [ˆb k + V bk, k û k = w k [ĉ ] w k k + V ck, ˆv k = w k w k [ ˆd k + V dk ],

6 Heght acceleraton Age Heght acceleraton Age Fgure : Curves measurng the heght acceleraton for 39 boys; (left) smoothed versons of raw observatons, (rght) automatcally algned curves. [ ] [ ] ˆβ k = w k ĉ ˆX ˆX k k k + Vxx w k ĉ k ˆX k(y ˆd k )+V xy V xcd, and ˆσ k = [ y w w k ĉ k ˆX k β kn ˆd k y Vx ˆβ k + ˆβ kvxx ˆβ k +ˆβ kvxcd + n V dk ], where ˆX k = â k x ˆb k, and Vxx, Vx, Vxcd are specal varance matrces whose components are functons of the posteror expectatons of Φ calculated n the E-step (the exact forms of these matrces can be found n []). 3 Expermental results and conclusons The results of a smple demonstraton of EM-based algnment (usng splnes and the learnng algorthm of the prevous secton, but wth no clusterng) are shown n Fgure. In the left plot are a set of smoothed curves representng the acceleraton of heght for each of 39 boys whose heghts were measured at 9 observaton tmes over the ages of to 8 []. Notce that the curves share a smlar shape but seem to be msalgned n tme due to ndvdual growth dynamcs. The rght plot shows the same acceleraton curves after processng from our splne algnment model (allowng only for translatons n tme). The x-axs n ths plot can be seen as canoncal (or average ) age. The algned curves n the rght plot of Fgure represent the average behavor n a much clearer way. For example, t appears there s an nterval of.5 years from peak (age.5) to trough (age 5) that descrbes the average cycle that all boys go through. The results demonstrate that t s common for mportant features of curves to be randomly translated n tme and that t s possble to use the data to recover these underlyng hdden transformatons usng our algnment models. Next we brefly present an applcaton of the jont clusterng-algnment model to the problem of gene expresson clusterng. We analyze the alpha arrest data descrbed n [3] that captures gene expresson levels at 7 mnute ntervals for two consecutve cell cycles (totalng 7 measurements per gene). Clusterng s often used n gene expresson analyss to reveal groups of genes wth smlar profles that may be physcally related to the same underlyng bologcal process (e.g., [3]). It s well-known that tme-delays play an mportant role n gene regulaton, and thus, curves measured over tme whch represent the same process may often be msalgned from each other. [4].

7 Expresson Expresson 5 5 Canoncal tme 5 5 Tme Expresson Expresson 5 5 Canoncal tme 5 5 Tme Expresson Expresson 5 5 Canoncal tme 5 5 Tme Fgure 3: Three clusters for the tme translaton algnment model (left) and the nonalgnment model (rght). Snce these gene expresson data are already normalzed, we dd not allow for transformatons n measurement space. We only allow for translatons n tme snce experts do not expect scalng n tme to be a factor n these data. Due to lmted space, we present a sngle case of comparson between a standard splne regresson mxture model (SRM) and an SRM that jontly allows for tme translatons. Ten random starts of EM were allowed for each algorthm wth the hghest lkelhood model selected for comparson for each algorthm. It s common to assume that there are fve dstnct clusters of genes n these data; as such we set K =5for each algorthm [3]. Three of the resultng clusters from the two methods are shown n Fgure 3. The left column of the fgure shows the output from the jont clusterng-algnment model, whle the rght column shows the output from the standard cluster model. It s mmedately obvous that the tme-algned clusters represent the mean behavor n a much clearer way. The overall cluster varance s much lower than n the non-algned clusterng. The results also demonstrate the appearance of cluster-dependent algnment effects. For example, the frst two clusters show large wthn-cluster msalgnment, whereas the thrd cluster does not. Interestngly, the actual clusterng s consderably dfferent between the two methods. In fact, only 57% of the expresson profles are assgned to common clusters between the two methods. Out-of-sample experments (not shown here) show that the jont model produces better predctve models than the standard clusterng method. Expermental results on a varety of other data sets are provded n [], ncludng applcatons to clusterng of cyclone trajectores.

8 4 Conclusons We proposed a general probablstc framework for jont clusterng and algnment of sets of curves. The expermental results ndcate that the approach provdes a new and useful tool for curve analyss n the face of underlyng hdden transformatons. The resultng EM-based learnng algorthms have tme-complexty that s lnear n the number of measurements n contrast, many exstng curve algnment algorthms themselves are O(n ) (e.g., dynamc tme warpng) wthout regard to clusterng. The ncorporaton of splnes gves the method an overall non-parametrc freedom whch leads to general applcablty. References [] J.O. Ramsay and B. W. Slverman. Functonal Data Analyss. Sprnger-Verlag, New York, NY, 997. [] Scott J. Gaffney. Probablstc Curve-Algned Clusterng and Predcton wth Regresson Mxture Models. Ph.D. Dssertaton, Unversty of Calforna, Irvne, 4. [3] Z. Bar-Joseph et al. A new approach to analyzng gene expresson tme seres data. Journal of Computatonal Bology, (3):34 356, 3. [4] B. J. Frey and N. Jojc. Transformaton-nvarant clusterng usng the EM algorthm. IEEE Trans. PAMI, 5(): 7, January 3. [5] H. Chu, J. Zhang, and A. Rangarajan. Unsupervsed learnng of an atlas from unlabeled pontsets. IEEE Trans. PAMI, 6():6 7, February 4. [6] A. D. J. Cross and E. R. Hancock. Graph matchng wth a dual-step EM algorthm. IEEE Trans. PAMI, ():36 53, November 998. [7] S. J. Gaffney and P. Smyth. Curve clusterng wth random effects regresson mxtures. In C. M. Bshop and B. J. Frey, edtors, Proc. Nnth Inter. Workshop on Artfcal Intellgence and Stats, Key West, FL, January [8] D. Chudova, S. J. Gaffney, and P. J. Smyth. Probablstc models for jont clusterng and tmewarpng of mult-dmensonal curves. In Proc. of the Nneteenth Conference on Uncertanty n Artfcal Intellgence (UAI-3), Acapulco, Mexco, August 7, 3. [9] D. Chudova, S. J. Gaffney, E. Mjolsness, and P. J. Smyth. Translaton-nvarant mxture models for curve clusterng. In Proc. Nnth ACM SIGKDD Inter. Conf. on Knowledge Dscovery and Data Mnng, Washngton D.C., August 4 7, New York, 3. ACM Press. [] S. Gaffney and P. Smyth. Trajectory clusterng wth mxtures of regresson models. In Surajt Chaudhur and Davd Madgan, edtors, Proc. Ffth ACM SIGKDD Inter. Conf. on Knowledge Dscovery and Data Mnng, August 5 8, pages 63 7, N.Y., 999. ACM Press. [] P. H. C. Elers. and B. D. Marx. Flexble smoothng wth B-splnes and penaltes. Statstcal Scence, ():89, 996. [] A. Gelman, J. B. Carln, H. S. Stern, and D. B. Rubn. Bayesan Data Analyss. Chapman & Hall, New York, NY, 995. [3] P. T. Spellman et al. Comprehensve dentfcaton of cell cycle-regulated genes of the yeast Saccharomyces cerevsae by mcroarray hybrdzaton. Molec. Bo. Cell, 9(): , December 998. [4] J. Aach and G. M. Church. Algnng gene expresson tme seres wth tme warpng algorthms. Bonformatcs, 7(6):495 58,.

Hermite Splines in Lie Groups as Products of Geodesics

Hermite Splines in Lie Groups as Products of Geodesics Hermte Splnes n Le Groups as Products of Geodescs Ethan Eade Updated May 28, 2017 1 Introducton 1.1 Goal Ths document defnes a curve n the Le group G parametrzed by tme and by structural parameters n the

More information

S1 Note. Basis functions.

S1 Note. Basis functions. S1 Note. Bass functons. Contents Types of bass functons...1 The Fourer bass...2 B-splne bass...3 Power and type I error rates wth dfferent numbers of bass functons...4 Table S1. Smulaton results of type

More information

Cluster Analysis of Electrical Behavior

Cluster Analysis of Electrical Behavior Journal of Computer and Communcatons, 205, 3, 88-93 Publshed Onlne May 205 n ScRes. http://www.scrp.org/ournal/cc http://dx.do.org/0.4236/cc.205.350 Cluster Analyss of Electrcal Behavor Ln Lu Ln Lu, School

More information

Analysis of Continuous Beams in General

Analysis of Continuous Beams in General Analyss of Contnuous Beams n General Contnuous beams consdered here are prsmatc, rgdly connected to each beam segment and supported at varous ponts along the beam. onts are selected at ponts of support,

More information

Feature Reduction and Selection

Feature Reduction and Selection Feature Reducton and Selecton Dr. Shuang LIANG School of Software Engneerng TongJ Unversty Fall, 2012 Today s Topcs Introducton Problems of Dmensonalty Feature Reducton Statstc methods Prncpal Components

More information

Lecture 5: Multilayer Perceptrons

Lecture 5: Multilayer Perceptrons Lecture 5: Multlayer Perceptrons Roger Grosse 1 Introducton So far, we ve only talked about lnear models: lnear regresson and lnear bnary classfers. We noted that there are functons that can t be represented

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

Learning the Kernel Parameters in Kernel Minimum Distance Classifier Learnng the Kernel Parameters n Kernel Mnmum Dstance Classfer Daoqang Zhang 1,, Songcan Chen and Zh-Hua Zhou 1* 1 Natonal Laboratory for Novel Software Technology Nanjng Unversty, Nanjng 193, Chna Department

More information

A Bilinear Model for Sparse Coding

A Bilinear Model for Sparse Coding A Blnear Model for Sparse Codng Davd B. Grmes and Rajesh P. N. Rao Department of Computer Scence and Engneerng Unversty of Washngton Seattle, WA 98195-2350, U.S.A. grmes,rao @cs.washngton.edu Abstract

More information

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics

NAG Fortran Library Chapter Introduction. G10 Smoothing in Statistics Introducton G10 NAG Fortran Lbrary Chapter Introducton G10 Smoothng n Statstcs Contents 1 Scope of the Chapter... 2 2 Background to the Problems... 2 2.1 Smoothng Methods... 2 2.2 Smoothng Splnes and Regresson

More information

CS 534: Computer Vision Model Fitting

CS 534: Computer Vision Model Fitting CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust

More information

Three supervised learning methods on pen digits character recognition dataset

Three supervised learning methods on pen digits character recognition dataset Three supervsed learnng methods on pen dgts character recognton dataset Chrs Flezach Department of Computer Scence and Engneerng Unversty of Calforna, San Dego San Dego, CA 92093 cflezac@cs.ucsd.edu Satoru

More information

Determining the Optimal Bandwidth Based on Multi-criterion Fusion

Determining the Optimal Bandwidth Based on Multi-criterion Fusion Proceedngs of 01 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 5 (01) (01) IACSIT Press, Sngapore Determnng the Optmal Bandwdth Based on Mult-crteron Fuson Ha-L Lang 1+, Xan-Mn

More information

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points; Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features

More information

Smoothing Spline ANOVA for variable screening

Smoothing Spline ANOVA for variable screening Smoothng Splne ANOVA for varable screenng a useful tool for metamodels tranng and mult-objectve optmzaton L. Rcco, E. Rgon, A. Turco Outlne RSM Introducton Possble couplng Test case MOO MOO wth Game Theory

More information

Mathematics 256 a course in differential equations for engineering students

Mathematics 256 a course in differential equations for engineering students Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the

More information

A New Approach For the Ranking of Fuzzy Sets With Different Heights

A New Approach For the Ranking of Fuzzy Sets With Different Heights New pproach For the ankng of Fuzzy Sets Wth Dfferent Heghts Pushpnder Sngh School of Mathematcs Computer pplcatons Thapar Unversty, Patala-7 00 Inda pushpndersnl@gmalcom STCT ankng of fuzzy sets plays

More information

Modeling Waveform Shapes with Random Effects Segmental Hidden Markov Models

Modeling Waveform Shapes with Random Effects Segmental Hidden Markov Models Modelng Waveform Shapes wth Random Effects Segmental Hdden Markov Models Seyoung Km, Padhrac Smyth Department of Computer Scence Unversty of Calforna, Irvne CA 9697-345 {sykm,smyth}@cs.uc.edu Abstract

More information

Lecture 4: Principal components

Lecture 4: Principal components /3/6 Lecture 4: Prncpal components 3..6 Multvarate lnear regresson MLR s optmal for the estmaton data...but poor for handlng collnear data Covarance matrx s not nvertble (large condton number) Robustness

More information

EXTENDED BIC CRITERION FOR MODEL SELECTION

EXTENDED BIC CRITERION FOR MODEL SELECTION IDIAP RESEARCH REPORT EXTEDED BIC CRITERIO FOR ODEL SELECTIO Itshak Lapdot Andrew orrs IDIAP-RR-0-4 Dalle olle Insttute for Perceptual Artfcal Intellgence P.O.Box 59 artgny Valas Swtzerland phone +4 7

More information

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization

Problem Definitions and Evaluation Criteria for Computational Expensive Optimization Problem efntons and Evaluaton Crtera for Computatonal Expensve Optmzaton B. Lu 1, Q. Chen and Q. Zhang 3, J. J. Lang 4, P. N. Suganthan, B. Y. Qu 6 1 epartment of Computng, Glyndwr Unversty, UK Faclty

More information

Adaptive Transfer Learning

Adaptive Transfer Learning Adaptve Transfer Learnng Bn Cao, Snno Jaln Pan, Yu Zhang, Dt-Yan Yeung, Qang Yang Hong Kong Unversty of Scence and Technology Clear Water Bay, Kowloon, Hong Kong {caobn,snnopan,zhangyu,dyyeung,qyang}@cse.ust.hk

More information

X- Chart Using ANOM Approach

X- Chart Using ANOM Approach ISSN 1684-8403 Journal of Statstcs Volume 17, 010, pp. 3-3 Abstract X- Chart Usng ANOM Approach Gullapall Chakravarth 1 and Chaluvad Venkateswara Rao Control lmts for ndvdual measurements (X) chart are

More information

APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET

APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET Jae-young Lee, Shahram Payandeh, and Ljljana Trajovć School of Engneerng Scence Smon Fraser Unversty 8888 Unversty

More information

Unsupervised Learning

Unsupervised Learning Pattern Recognton Lecture 8 Outlne Introducton Unsupervsed Learnng Parametrc VS Non-Parametrc Approach Mxture of Denstes Maxmum-Lkelhood Estmates Clusterng Prof. Danel Yeung School of Computer Scence and

More information

Lecture 5: Probability Distributions. Random Variables

Lecture 5: Probability Distributions. Random Variables Lecture 5: Probablty Dstrbutons Random Varables Probablty Dstrbutons Dscrete Random Varables Contnuous Random Varables and ther Dstrbutons Dscrete Jont Dstrbutons Contnuous Jont Dstrbutons Independent

More information

APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET

APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET APPLICATION OF PREDICTION-BASED PARTICLE FILTERS FOR TELEOPERATIONS OVER THE INTERNET Jae-young Lee, Shahram Payandeh, and Ljljana Trajovć School of Engneerng Scence Smon Fraser Unversty 8888 Unversty

More information

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning Outlne Artfcal Intellgence and ts applcatons Lecture 8 Unsupervsed Learnng Professor Danel Yeung danyeung@eee.org Dr. Patrck Chan patrckchan@eee.org South Chna Unversty of Technology, Chna Introducton

More information

y and the total sum of

y and the total sum of Lnear regresson Testng for non-lnearty In analytcal chemstry, lnear regresson s commonly used n the constructon of calbraton functons requred for analytcal technques such as gas chromatography, atomc absorpton

More information

A Binarization Algorithm specialized on Document Images and Photos

A Binarization Algorithm specialized on Document Images and Photos A Bnarzaton Algorthm specalzed on Document mages and Photos Ergna Kavalleratou Dept. of nformaton and Communcaton Systems Engneerng Unversty of the Aegean kavalleratou@aegean.gr Abstract n ths paper, a

More information

Biostatistics 615/815

Biostatistics 615/815 The E-M Algorthm Bostatstcs 615/815 Lecture 17 Last Lecture: The Smplex Method General method for optmzaton Makes few assumptons about functon Crawls towards mnmum Some recommendatons Multple startng ponts

More information

Hybridization of Expectation-Maximization and K-Means Algorithms for Better Clustering Performance

Hybridization of Expectation-Maximization and K-Means Algorithms for Better Clustering Performance BULGARIAN ACADEMY OF SCIENCES CYBERNETICS AND INFORMATION TECHNOLOGIES Volume 16, No 2 Sofa 2016 Prnt ISSN: 1311-9702; Onlne ISSN: 1314-4081 DOI: 10.1515/cat-2016-0017 Hybrdzaton of Expectaton-Maxmzaton

More information

A Robust Method for Estimating the Fundamental Matrix

A Robust Method for Estimating the Fundamental Matrix Proc. VIIth Dgtal Image Computng: Technques and Applcatons, Sun C., Talbot H., Ourseln S. and Adraansen T. (Eds.), 0- Dec. 003, Sydney A Robust Method for Estmatng the Fundamental Matrx C.L. Feng and Y.S.

More information

User Authentication Based On Behavioral Mouse Dynamics Biometrics

User Authentication Based On Behavioral Mouse Dynamics Biometrics User Authentcaton Based On Behavoral Mouse Dynamcs Bometrcs Chee-Hyung Yoon Danel Donghyun Km Department of Computer Scence Department of Computer Scence Stanford Unversty Stanford Unversty Stanford, CA

More information

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,

More information

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images

A B-Snake Model Using Statistical and Geometric Information - Applications to Medical Images A B-Snake Model Usng Statstcal and Geometrc Informaton - Applcatons to Medcal Images Yue Wang, Eam Khwang Teoh and Dnggang Shen 2 School of Electrcal and Electronc Engneerng, Nanyang Technologcal Unversty

More information

EECS 730 Introduction to Bioinformatics Sequence Alignment. Luke Huan Electrical Engineering and Computer Science

EECS 730 Introduction to Bioinformatics Sequence Alignment. Luke Huan Electrical Engineering and Computer Science EECS 730 Introducton to Bonformatcs Sequence Algnment Luke Huan Electrcal Engneerng and Computer Scence http://people.eecs.ku.edu/~huan/ HMM Π s a set of states Transton Probabltes a kl Pr( l 1 k Probablty

More information

Support Vector Machines

Support Vector Machines Support Vector Machnes Decson surface s a hyperplane (lne n 2D) n feature space (smlar to the Perceptron) Arguably, the most mportant recent dscovery n machne learnng In a nutshell: map the data to a predetermned

More information

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance

Tsinghua University at TAC 2009: Summarizing Multi-documents by Information Distance Tsnghua Unversty at TAC 2009: Summarzng Mult-documents by Informaton Dstance Chong Long, Mnle Huang, Xaoyan Zhu State Key Laboratory of Intellgent Technology and Systems, Tsnghua Natonal Laboratory for

More information

Wavefront Reconstructor

Wavefront Reconstructor A Dstrbuted Smplex B-Splne Based Wavefront Reconstructor Coen de Vsser and Mchel Verhaegen 14-12-201212 2012 Delft Unversty of Technology Contents Introducton Wavefront reconstructon usng Smplex B-Splnes

More information

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices

An Application of the Dulmage-Mendelsohn Decomposition to Sparse Null Space Bases of Full Row Rank Matrices Internatonal Mathematcal Forum, Vol 7, 2012, no 52, 2549-2554 An Applcaton of the Dulmage-Mendelsohn Decomposton to Sparse Null Space Bases of Full Row Rank Matrces Mostafa Khorramzadeh Department of Mathematcal

More information

Empirical Distributions of Parameter Estimates. in Binary Logistic Regression Using Bootstrap

Empirical Distributions of Parameter Estimates. in Binary Logistic Regression Using Bootstrap Int. Journal of Math. Analyss, Vol. 8, 4, no. 5, 7-7 HIKARI Ltd, www.m-hkar.com http://dx.do.org/.988/jma.4.494 Emprcal Dstrbutons of Parameter Estmates n Bnary Logstc Regresson Usng Bootstrap Anwar Ftranto*

More information

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data Malaysan Journal of Mathematcal Scences 11(S) Aprl : 35 46 (2017) Specal Issue: The 2nd Internatonal Conference and Workshop on Mathematcal Analyss (ICWOMA 2016) MALAYSIAN JOURNAL OF MATHEMATICAL SCIENCES

More information

Mixed Linear System Estimation and Identification

Mixed Linear System Estimation and Identification 48th IEEE Conference on Decson and Control, Shangha, Chna, December 2009 Mxed Lnear System Estmaton and Identfcaton A. Zymns S. Boyd D. Gornevsky Abstract We consder a mxed lnear system model, wth both

More information

Fast Sparse Gaussian Processes Learning for Man-Made Structure Classification

Fast Sparse Gaussian Processes Learning for Man-Made Structure Classification Fast Sparse Gaussan Processes Learnng for Man-Made Structure Classfcaton Hang Zhou Insttute for Vson Systems Engneerng, Dept Elec. & Comp. Syst. Eng. PO Box 35, Monash Unversty, Clayton, VIC 3800, Australa

More information

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr)

Helsinki University Of Technology, Systems Analysis Laboratory Mat Independent research projects in applied mathematics (3 cr) Helsnk Unversty Of Technology, Systems Analyss Laboratory Mat-2.08 Independent research projects n appled mathematcs (3 cr) "! #$&% Antt Laukkanen 506 R ajlaukka@cc.hut.f 2 Introducton...3 2 Multattrbute

More information

For instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1)

For instance, ; the five basic number-sets are increasingly more n A B & B A A = B (1) Secton 1.2 Subsets and the Boolean operatons on sets If every element of the set A s an element of the set B, we say that A s a subset of B, or that A s contaned n B, or that B contans A, and we wrte A

More information

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching

A Fast Visual Tracking Algorithm Based on Circle Pixels Matching A Fast Vsual Trackng Algorthm Based on Crcle Pxels Matchng Zhqang Hou hou_zhq@sohu.com Chongzhao Han czhan@mal.xjtu.edu.cn Ln Zheng Abstract: A fast vsual trackng algorthm based on crcle pxels matchng

More information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS ARPN Journal of Engneerng and Appled Scences 006-017 Asan Research Publshng Network (ARPN). All rghts reserved. NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS Igor Grgoryev, Svetlana

More information

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision

SLAM Summer School 2006 Practical 2: SLAM using Monocular Vision SLAM Summer School 2006 Practcal 2: SLAM usng Monocular Vson Javer Cvera, Unversty of Zaragoza Andrew J. Davson, Imperal College London J.M.M Montel, Unversty of Zaragoza. josemar@unzar.es, jcvera@unzar.es,

More information

Edge Detection in Noisy Images Using the Support Vector Machines

Edge Detection in Noisy Images Using the Support Vector Machines Edge Detecton n Nosy Images Usng the Support Vector Machnes Hlaro Gómez-Moreno, Saturnno Maldonado-Bascón, Francsco López-Ferreras Sgnal Theory and Communcatons Department. Unversty of Alcalá Crta. Madrd-Barcelona

More information

Face Recognition University at Buffalo CSE666 Lecture Slides Resources:

Face Recognition University at Buffalo CSE666 Lecture Slides Resources: Face Recognton Unversty at Buffalo CSE666 Lecture Sldes Resources: http://www.face-rec.org/algorthms/ Overvew of face recognton algorthms Correlaton - Pxel based correspondence between two face mages Structural

More information

High-Boost Mesh Filtering for 3-D Shape Enhancement

High-Boost Mesh Filtering for 3-D Shape Enhancement Hgh-Boost Mesh Flterng for 3-D Shape Enhancement Hrokazu Yagou Λ Alexander Belyaev y Damng We z Λ y z ; ; Shape Modelng Laboratory, Unversty of Azu, Azu-Wakamatsu 965-8580 Japan y Computer Graphcs Group,

More information

Understanding K-Means Non-hierarchical Clustering

Understanding K-Means Non-hierarchical Clustering SUNY Albany - Techncal Report 0- Understandng K-Means Non-herarchcal Clusterng Ian Davdson State Unversty of New York, 1400 Washngton Ave., Albany, 105. DAVIDSON@CS.ALBANY.EDU Abstract The K-means algorthm

More information

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Why consder unlabeled samples?. Collectng and labelng large set of samples s costly Gettng recorded speech s free, labelng s tme consumng 2. Classfer could be desgned

More information

TN348: Openlab Module - Colocalization

TN348: Openlab Module - Colocalization TN348: Openlab Module - Colocalzaton Topc The Colocalzaton module provdes the faclty to vsualze and quantfy colocalzaton between pars of mages. The Colocalzaton wndow contans a prevew of the two mages

More information

Parameter estimation for incomplete bivariate longitudinal data in clinical trials

Parameter estimation for incomplete bivariate longitudinal data in clinical trials Parameter estmaton for ncomplete bvarate longtudnal data n clncal trals Naum M. Khutoryansky Novo Nordsk Pharmaceutcals, Inc., Prnceton, NJ ABSTRACT Bvarate models are useful when analyzng longtudnal data

More information

Classifier Selection Based on Data Complexity Measures *

Classifier Selection Based on Data Complexity Measures * Classfer Selecton Based on Data Complexty Measures * Edth Hernández-Reyes, J.A. Carrasco-Ochoa, and J.Fco. Martínez-Trndad Natonal Insttute for Astrophyscs, Optcs and Electroncs, Lus Enrque Erro No.1 Sta.

More information

Synthesizer 1.0. User s Guide. A Varying Coefficient Meta. nalytic Tool. Z. Krizan Employing Microsoft Excel 2007

Synthesizer 1.0. User s Guide. A Varying Coefficient Meta. nalytic Tool. Z. Krizan Employing Microsoft Excel 2007 Syntheszer 1.0 A Varyng Coeffcent Meta Meta-Analytc nalytc Tool Employng Mcrosoft Excel 007.38.17.5 User s Gude Z. Krzan 009 Table of Contents 1. Introducton and Acknowledgments 3. Operatonal Functons

More information

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour

6.854 Advanced Algorithms Petar Maymounkov Problem Set 11 (November 23, 2005) With: Benjamin Rossman, Oren Weimann, and Pouya Kheradpour 6.854 Advanced Algorthms Petar Maymounkov Problem Set 11 (November 23, 2005) Wth: Benjamn Rossman, Oren Wemann, and Pouya Kheradpour Problem 1. We reduce vertex cover to MAX-SAT wth weghts, such that the

More information

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers

Content Based Image Retrieval Using 2-D Discrete Wavelet with Texture Feature with Different Classifiers IOSR Journal of Electroncs and Communcaton Engneerng (IOSR-JECE) e-issn: 78-834,p- ISSN: 78-8735.Volume 9, Issue, Ver. IV (Mar - Apr. 04), PP 0-07 Content Based Image Retreval Usng -D Dscrete Wavelet wth

More information

Clustering Dynamic Textures with the Hierarchical EM Algorithm for Modeling Video

Clustering Dynamic Textures with the Hierarchical EM Algorithm for Modeling Video IEEE TRANS. ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE, TO APPEAR, 2013 1 Clusterng Dynamc Textures wth the Herarchcal EM Algorthm for Modelng Vdeo Adeel Mumtaz, Emanuele Covello, Gert. R. G. Lanckret,

More information

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters

Proper Choice of Data Used for the Estimation of Datum Transformation Parameters Proper Choce of Data Used for the Estmaton of Datum Transformaton Parameters Hakan S. KUTOGLU, Turkey Key words: Coordnate systems; transformaton; estmaton, relablty. SUMMARY Advances n technologes and

More information

Fusion Performance Model for Distributed Tracking and Classification

Fusion Performance Model for Distributed Tracking and Classification Fuson Performance Model for Dstrbuted rackng and Classfcaton K.C. Chang and Yng Song Dept. of SEOR, School of I&E George Mason Unversty FAIRFAX, VA kchang@gmu.edu Martn Lggns Verdan Systems Dvson, Inc.

More information

The Codesign Challenge

The Codesign Challenge ECE 4530 Codesgn Challenge Fall 2007 Hardware/Software Codesgn The Codesgn Challenge Objectves In the codesgn challenge, your task s to accelerate a gven software reference mplementaton as fast as possble.

More information

High Dimensional Data Clustering

High Dimensional Data Clustering Hgh Dmensonal Data Clusterng Charles Bouveyron 1,2, Stéphane Grard 1, and Cordela Schmd 2 1 LMC-IMAG, BP 53, Unversté Grenoble 1, 38041 Grenoble Cede 9, France charles.bouveyron@mag.fr, stephane.grard@mag.fr

More information

Support Vector Machines

Support Vector Machines /9/207 MIST.6060 Busness Intellgence and Data Mnng What are Support Vector Machnes? Support Vector Machnes Support Vector Machnes (SVMs) are supervsed learnng technques that analyze data and recognze patterns.

More information

An Accurate Evaluation of Integrals in Convex and Non convex Polygonal Domain by Twelve Node Quadrilateral Finite Element Method

An Accurate Evaluation of Integrals in Convex and Non convex Polygonal Domain by Twelve Node Quadrilateral Finite Element Method Internatonal Journal of Computatonal and Appled Mathematcs. ISSN 89-4966 Volume, Number (07), pp. 33-4 Research Inda Publcatons http://www.rpublcaton.com An Accurate Evaluaton of Integrals n Convex and

More information

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz

Compiler Design. Spring Register Allocation. Sample Exercises and Solutions. Prof. Pedro C. Diniz Compler Desgn Sprng 2014 Regster Allocaton Sample Exercses and Solutons Prof. Pedro C. Dnz USC / Informaton Scences Insttute 4676 Admralty Way, Sute 1001 Marna del Rey, Calforna 90292 pedro@s.edu Regster

More information

Recognizing Faces. Outline

Recognizing Faces. Outline Recognzng Faces Drk Colbry Outlne Introducton and Motvaton Defnng a feature vector Prncpal Component Analyss Lnear Dscrmnate Analyss !"" #$""% http://www.nfotech.oulu.f/annual/2004 + &'()*) '+)* 2 ! &

More information

Learning-Based Top-N Selection Query Evaluation over Relational Databases

Learning-Based Top-N Selection Query Evaluation over Relational Databases Learnng-Based Top-N Selecton Query Evaluaton over Relatonal Databases Lang Zhu *, Wey Meng ** * School of Mathematcs and Computer Scence, Hebe Unversty, Baodng, Hebe 071002, Chna, zhu@mal.hbu.edu.cn **

More information

SENSITIVITY ANALYSIS IN LINEAR PROGRAMMING USING A CALCULATOR

SENSITIVITY ANALYSIS IN LINEAR PROGRAMMING USING A CALCULATOR SENSITIVITY ANALYSIS IN LINEAR PROGRAMMING USING A CALCULATOR Judth Aronow Rchard Jarvnen Independent Consultant Dept of Math/Stat 559 Frost Wnona State Unversty Beaumont, TX 7776 Wnona, MN 55987 aronowju@hal.lamar.edu

More information

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009.

Assignment # 2. Farrukh Jabeen Algorithms 510 Assignment #2 Due Date: June 15, 2009. Farrukh Jabeen Algorthms 51 Assgnment #2 Due Date: June 15, 29. Assgnment # 2 Chapter 3 Dscrete Fourer Transforms Implement the FFT for the DFT. Descrbed n sectons 3.1 and 3.2. Delverables: 1. Concse descrpton

More information

Hierarchical clustering for gene expression data analysis

Hierarchical clustering for gene expression data analysis Herarchcal clusterng for gene expresson data analyss Gorgo Valentn e-mal: valentn@ds.unm.t Clusterng of Mcroarray Data. Clusterng of gene expresson profles (rows) => dscovery of co-regulated and functonally

More information

Classification / Regression Support Vector Machines

Classification / Regression Support Vector Machines Classfcaton / Regresson Support Vector Machnes Jeff Howbert Introducton to Machne Learnng Wnter 04 Topcs SVM classfers for lnearly separable classes SVM classfers for non-lnearly separable classes SVM

More information

Modeling Inter-cluster and Intra-cluster Discrimination Among Triphones

Modeling Inter-cluster and Intra-cluster Discrimination Among Triphones Modelng Inter-cluster and Intra-cluster Dscrmnaton Among Trphones Tom Ko, Bran Mak and Dongpeng Chen Department of Computer Scence and Engneerng The Hong Kong Unversty of Scence and Technology Clear Water

More information

Context-Specific Bayesian Clustering for Gene Expression Data

Context-Specific Bayesian Clustering for Gene Expression Data Context-Specfc Bayesan Clusterng for Gene Expresson Data Yoseph Barash School of Computer Scence & Engneerng Hebrew Unversty, Jerusalem, 91904, Israel hoan@cs.huj.ac.l Nr Fredman School of Computer Scence

More information

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval

Fuzzy C-Means Initialized by Fixed Threshold Clustering for Improving Image Retrieval Fuzzy -Means Intalzed by Fxed Threshold lusterng for Improvng Image Retreval NAWARA HANSIRI, SIRIPORN SUPRATID,HOM KIMPAN 3 Faculty of Informaton Technology Rangst Unversty Muang-Ake, Paholyotn Road, Patumtan,

More information

On Supporting Identification in a Hand-Based Biometric Framework

On Supporting Identification in a Hand-Based Biometric Framework On Supportng Identfcaton n a Hand-Based Bometrc Framework Pe-Fang Guo 1, Prabr Bhattacharya 2, and Nawwaf Kharma 1 1 Electrcal & Computer Engneerng, Concorda Unversty, 1455 de Masonneuve Blvd., Montreal,

More information

GSLM Operations Research II Fall 13/14

GSLM Operations Research II Fall 13/14 GSLM 58 Operatons Research II Fall /4 6. Separable Programmng Consder a general NLP mn f(x) s.t. g j (x) b j j =. m. Defnton 6.. The NLP s a separable program f ts objectve functon and all constrants are

More information

LECTURE : MANIFOLD LEARNING

LECTURE : MANIFOLD LEARNING LECTURE : MANIFOLD LEARNING Rta Osadchy Some sldes are due to L.Saul, V. C. Raykar, N. Verma Topcs PCA MDS IsoMap LLE EgenMaps Done! Dmensonalty Reducton Data representaton Inputs are real-valued vectors

More information

Journal of Chemical and Pharmaceutical Research, 2014, 6(6): Research Article

Journal of Chemical and Pharmaceutical Research, 2014, 6(6): Research Article Avalable onlne www.jocpr.com Journal of Chemcal and Pharmaceutcal Research, 2014, 6(6):2512-2520 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCPRC5 Communty detecton model based on ncremental EM clusterng

More information

CS434a/541a: Pattern Recognition Prof. Olga Veksler. Lecture 15

CS434a/541a: Pattern Recognition Prof. Olga Veksler. Lecture 15 CS434a/541a: Pattern Recognton Prof. Olga Veksler Lecture 15 Today New Topc: Unsupervsed Learnng Supervsed vs. unsupervsed learnng Unsupervsed learnng Net Tme: parametrc unsupervsed learnng Today: nonparametrc

More information

TESTING AND IMPROVING LOCAL ADAPTIVE IMPORTANCE SAMPLING IN LJF LOCAL-JT IN MULTIPLY SECTIONED BAYESIAN NETWORKS

TESTING AND IMPROVING LOCAL ADAPTIVE IMPORTANCE SAMPLING IN LJF LOCAL-JT IN MULTIPLY SECTIONED BAYESIAN NETWORKS TESTING AND IMPROVING LOCAL ADAPTIVE IMPORTANCE SAMPLING IN LJF LOCAL-JT IN MULTIPLY SECTIONED BAYESIAN NETWORKS Dan Wu 1 and Sona Bhatt 2 1 School of Computer Scence Unversty of Wndsor, Wndsor, Ontaro

More information

Econometrics 2. Panel Data Methods. Advanced Panel Data Methods I

Econometrics 2. Panel Data Methods. Advanced Panel Data Methods I Panel Data Methods Econometrcs 2 Advanced Panel Data Methods I Last tme: Panel data concepts and the two-perod case (13.3-4) Unobserved effects model: Tme-nvarant and dosyncratc effects Omted varables

More information

Parallelism for Nested Loops with Non-uniform and Flow Dependences

Parallelism for Nested Loops with Non-uniform and Flow Dependences Parallelsm for Nested Loops wth Non-unform and Flow Dependences Sam-Jn Jeong Dept. of Informaton & Communcaton Engneerng, Cheonan Unversty, 5, Anseo-dong, Cheonan, Chungnam, 330-80, Korea. seong@cheonan.ac.kr

More information

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur

FEATURE EXTRACTION. Dr. K.Vijayarekha. Associate Dean School of Electrical and Electronics Engineering SASTRA University, Thanjavur FEATURE EXTRACTION Dr. K.Vjayarekha Assocate Dean School of Electrcal and Electroncs Engneerng SASTRA Unversty, Thanjavur613 41 Jont Intatve of IITs and IISc Funded by MHRD Page 1 of 8 Table of Contents

More information

The Research of Support Vector Machine in Agricultural Data Classification

The Research of Support Vector Machine in Agricultural Data Classification The Research of Support Vector Machne n Agrcultural Data Classfcaton Le Sh, Qguo Duan, Xnmng Ma, Me Weng College of Informaton and Management Scence, HeNan Agrcultural Unversty, Zhengzhou 45000 Chna Zhengzhou

More information

Solving two-person zero-sum game by Matlab

Solving two-person zero-sum game by Matlab Appled Mechancs and Materals Onlne: 2011-02-02 ISSN: 1662-7482, Vols. 50-51, pp 262-265 do:10.4028/www.scentfc.net/amm.50-51.262 2011 Trans Tech Publcatons, Swtzerland Solvng two-person zero-sum game by

More information

Classifying Acoustic Transient Signals Using Artificial Intelligence

Classifying Acoustic Transient Signals Using Artificial Intelligence Classfyng Acoustc Transent Sgnals Usng Artfcal Intellgence Steve Sutton, Unversty of North Carolna At Wlmngton (suttons@charter.net) Greg Huff, Unversty of North Carolna At Wlmngton (jgh7476@uncwl.edu)

More information

APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT

APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT 3. - 5. 5., Brno, Czech Republc, EU APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT Abstract Josef TOŠENOVSKÝ ) Lenka MONSPORTOVÁ ) Flp TOŠENOVSKÝ

More information

Incremental MQDF Learning for Writer Adaptive Handwriting Recognition 1

Incremental MQDF Learning for Writer Adaptive Handwriting Recognition 1 200 2th Internatonal Conference on Fronters n Handwrtng Recognton Incremental MQDF Learnng for Wrter Adaptve Handwrtng Recognton Ka Dng, Lanwen Jn * School of Electronc and Informaton Engneerng, South

More information

A Post Randomization Framework for Privacy-Preserving Bayesian. Network Parameter Learning

A Post Randomization Framework for Privacy-Preserving Bayesian. Network Parameter Learning A Post Randomzaton Framework for Prvacy-Preservng Bayesan Network Parameter Learnng JIANJIE MA K.SIVAKUMAR School Electrcal Engneerng and Computer Scence, Washngton State Unversty Pullman, WA. 9964-75

More information

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements

2x x l. Module 3: Element Properties Lecture 4: Lagrange and Serendipity Elements Module 3: Element Propertes Lecture : Lagrange and Serendpty Elements 5 In last lecture note, the nterpolaton functons are derved on the bass of assumed polynomal from Pascal s trangle for the fled varable.

More information

Lecture #15 Lecture Notes

Lecture #15 Lecture Notes Lecture #15 Lecture Notes The ocean water column s very much a 3-D spatal entt and we need to represent that structure n an economcal way to deal wth t n calculatons. We wll dscuss one way to do so, emprcal

More information

A User Selection Method in Advertising System

A User Selection Method in Advertising System Int. J. Communcatons, etwork and System Scences, 2010, 3, 54-58 do:10.4236/jcns.2010.31007 Publshed Onlne January 2010 (http://www.scrp.org/journal/jcns/). A User Selecton Method n Advertsng System Shy

More information

An Ensemble Learning algorithm for Blind Signal Separation Problem

An Ensemble Learning algorithm for Blind Signal Separation Problem An Ensemble Learnng algorthm for Blnd Sgnal Separaton Problem Yan L 1 and Peng Wen 1 Department of Mathematcs and Computng, Faculty of Engneerng and Surveyng The Unversty of Southern Queensland, Queensland,

More information

LEARNING A WARPED SUBSPACE MODEL OF FACES WITH IMAGES OF UNKNOWN POSE AND ILLUMINATION

LEARNING A WARPED SUBSPACE MODEL OF FACES WITH IMAGES OF UNKNOWN POSE AND ILLUMINATION LEARNING A WARPED SUBSPACE MODEL OF FACES WITH IMAGES OF UNKNOWN POSE AND ILLUMINATION Jhun Hamm, and Danel D. Lee GRASP Laboratory, Unversty of Pennsylvana, 3330 Walnut Street, Phladelpha, PA, USA jhham@seas.upenn.edu,

More information

Exercises (Part 4) Introduction to R UCLA/CCPR. John Fox, February 2005

Exercises (Part 4) Introduction to R UCLA/CCPR. John Fox, February 2005 Exercses (Part 4) Introducton to R UCLA/CCPR John Fox, February 2005 1. A challengng problem: Iterated weghted least squares (IWLS) s a standard method of fttng generalzed lnear models to data. As descrbed

More information

AVO Modeling of Monochromatic Spherical Waves: Comparison to Band-Limited Waves

AVO Modeling of Monochromatic Spherical Waves: Comparison to Band-Limited Waves AVO Modelng of Monochromatc Sphercal Waves: Comparson to Band-Lmted Waves Charles Ursenbach* Unversty of Calgary, Calgary, AB, Canada ursenbach@crewes.org and Arnm Haase Unversty of Calgary, Calgary, AB,

More information

An Optimal Algorithm for Prufer Codes *

An Optimal Algorithm for Prufer Codes * J. Software Engneerng & Applcatons, 2009, 2: 111-115 do:10.4236/jsea.2009.22016 Publshed Onlne July 2009 (www.scrp.org/journal/jsea) An Optmal Algorthm for Prufer Codes * Xaodong Wang 1, 2, Le Wang 3,

More information