Feature Extraction and Dimensionality Reduction in SVM Speaker Recognition

Size: px
Start display at page:

Download "Feature Extraction and Dimensionality Reduction in SVM Speaker Recognition"

Transcription

1 Feature Extracton and Dmensonalty Reducton n SVM Speaker Recognton hembsle Mazbuko, themb@crg.ee.uct.ac.za Danel J. Mashao, danel@ebe.uct.ac.za Department of Electrcal Engneerng, Unversty of Cape own Abstract he Support Vector Machne s a dscrmnatve classfer whch has acheved mpressve results n several pattern recognton tasks. Its applcablty s however lmted by the computatonal expense. he SVM tranng and testng tmes ncreases wth ncreasng amounts of data. A possble soluton s the reducton of the number of computatons wthn the SVM kernel by reducng the dmensonalty of the observaton vectors. In ths paper we apply ths concept of reducng the data dmensonalty to decrease the complexty to an SVM-based speaker verfcaton task, carred out on a subset of the IS 2000 speaker recognton evaluaton. he dmensonalty reducton s performed by applyng Prncpal Component Analyss to the feature vector. he results show degradaton n the performance of SVM when feature extracton s appled. here s however also a sgnfcant decrease n the tranng and testng tmes. Key erms speaker verfcaton, support vector machne, feature transformaton A UOMAIC speaker recognton s the task of usng machnes to recognze people usng speech as a bometrc. he speech nput s converted nto a feature vector representaton. Lnear Predctve Coeffcents (LPC) and Mel- Frequency Cepstral Coeffcents (MFCC) are popular feature sets. hese feature sets are used to create models whch represent each speaker. In the testng phase the nput speech s compared to the model and a classfcaton engne s used to decde who the test speaker s. he Support Vector Machne (SVM) s a dscrmnatve classfer whch has been successfully appled dfferent pattern recognton felds. In the speech processng feld the authors of [] appled SVM to a speaker and language recognton task on the IS 2003 database wth good results. In [2] the author appled SVM to a speaker verfcaton task and found that when combned wth a Gaussan Mxture Model (GMM) classfer, SVM mproved on the performance of the more popular classfer. he power of SVM les n the mplct transformaton of the nput space to a hgher dmensonal space. Because the transformaton s performed mplctly, there are no computatons explctly performed n the hgh dmensonal space. Intutvely, ths means that the dmenson of the nput dataset does not affect the performance of SVM. However, accordng to [3], the performance of SVM s dependent on the radus of the data, whch ncreases wth the number of features. A major hndrance n the applcaton of SVM s that the classfer suffers from extremely long tranng tme especally as the sze of the data, or number of observaton vectors, ncreases. Research studes have shown an mprovement n the accuracy of dscrmnatve classfers by applyng feature extracton technques [4,5]. he bass of these studes s that by applyng feature extracton we can remove redundances n the feature vectors by extractng and usng only those features whch are most relevant. Prncpal component analyss (PCA) s a relatvely old and well developed lnear ndependent feature extracton technque whch has been appled to several pattern recognton tasks. Pattern recognton tasks are dvded nto two phases; feature analyss and classfcaton. Feature extracton s a part of the feature analyss phase where we attempt to reduce redundancy n the feature vectors. An attracton of technques lkes PCA (others nclude Lnear Dscrmnant Analyss and Independent Component Analyss) s ther data dmensonalty reducton capabltes. In [5] the author proposed what he called the Reduced Dmensonal Support Vector Machne whch apples feature extracton and dmensonalty reducton technques. he concept behnd RDSVM s that the computatonal burden on SVM classfcaton can be decreased by reducng the number of computatons wthn the classfer. It s generally not possble nor desrable to reduce the number of observaton vectors as ths mght lead to loss of mportant nformaton. However, SVM performance suffers from long tranng and testng tmes when the number of observaton vectors s too hgh. So, RDSVM reduces the dmensonalty of the observaton vectors as a way of reducng the total number of SVM computatons. In the experments dscussed n ths paper, the effect of feature extracton on the system accuracy and speed of SVM tranng and classfcaton on a speaker verfcaton task was studed. he rest of ths paper s organzed as follows, Secton I gves an ntroducton to the SVM classfer. In Secton II we present the mathematcal formulaton of Prncpal Component Analyss. In Secton III a dscusson of speaker recognton s gven. Secton IV s the expermental setup and the results of the experments conducted. Secton V s a dscusson of the results and Secton VI s the concluson and a short dscusson of future work. I. SUPPOR VECOR MACHIE CLASSIFIER he Support Vector Machne s a powerful dscrmnatve classfer whch maps the nput onto a hgh dmensonal space d by Φ : R I ; and then fnds an optmal hyperplane to separate the data n that space. hs separatng hyperplane s found by maxmzng the dstance of the closest patterns [6]. he new space s often referred to as the feature space. Here we wll present a smplfed, general explanaton of SVM.

2 Suppose we have a bnary classfcaton problem as shown n fgure where each example belongs to ether class + or -. SVM seeks to maxmze the margn between the two classes by fndng the separatng hyperplane whch les halfway between the data classes. We can, wthout loss of generalty, consder the case of data that are non-lnearly separable. When the data are transformed by some non-lnear transformaton onto a hgher dmensonal space, they spread out allowng a separatng hyperplane to be found n the feature space. Fgure : Fndng the separatng hyperplane for non-lnearly separable data he support vectors are those data ponts that le on S and S 2, le wthn the margn, or are msclassfed. he hyperplane n the hgh dmensonal transform space results n a complex decson surface n the nput data space. SVM allows for these msclassfcatons n tranng wth a user-defned cost parameter C. hs way the msclassfcatons are lmted whle stll avodng over-fttng the tranng data [2,7]. In general the SVM s computed by usng the kernel trck so f ( x) = α y K( s,x) = b = where K s some kernel functon such that and s are the support vectors. a.) ( x,x ) = Φ( x ) Φ( x ) j j () K (2) SVM Characterstcs A dstngushng characterstc of SVM s ts strong foundaton on statstcal learnng theory whch establshes a bound on the generalzaton error (the error rate of a learnng machne on unseen data) thus mprovng the classfcaton results for unseen patterns [6]. he SVM mnmzes ths bound by maxmzng the margn. Also, snce SVM projects the nput onto a hgher dmenson space, the margn maxmzaton s ndependent of the orgnal dmenson of the nput space, thus SVM successfully avods the curse of dmensonalty whch some classfers suffer from. Determnng most approprate choce of kernel for a partcular task. Once a kernel has been chosen, there stll remans the ssue of optmzng the parameters of the kernel. For nstance, n [8] the authors conclude that there s an optmum C value for each dataset. he desgn of SVM s optmzed for bnary classfcaton whch can lmt ts applcablty to mult-class classfcaton tasks. Although the SVM successfully avods the curse of dmensonalty, the technque can be ncredbly slow, especally as the sze of the data ncreases. hs s true for both the tranng and test phases. here are stll several challenges whch need o be addressed n SVM applcatons. In [7] the author suggested that there mght be a possblty for lmtng the computatonal load of SVM by reducng the data dmensonalty whch would decrease the number of computatons that have to be performed wthn the SVM. he followng secton presents an overvew of a popular dmensonalty reducton technque. II. PRICIPAL COMPOE AALYSIS A possble approach to mprove the classfcaton performance of SVM s to operate the classfer n a feature space n whch the classes are nherently separated [7]. hs feature space s typcally a mult-dmensonal space resultng from transformng the nput space va some lnear or non-lnear transformaton. Prncpal Component Analyss (PCA) s one such transform. PCA s also referred to as the dscrete Karhunen-Loève ransform (KL) or the Hotellng transform. he central prncple n PCA s to transform the nput space onto a feature space where the data show maxmal varance. he followng s a bref dscusson of the mathematcal formulaton of PCA. he PCA formulaton gven here closely follows that of [2,5]. Let x be an m-dmensonal nput data vector such that x = ( x,,x )... m. We estmate the sample mean by µ ( x ) = (3) where s the number of samples. he sample covarance matrx C becomes b.) Challenges n SVM SVM s; however, not wthout ts lmtatons. Some of these S are: margn S 2 on-lnear transformaton on-lnearly separable n nput space Separatng hyperplane

3 C = = = = ( x µ ) ( x ) µ x x µµ (4) (5) Input speech Parameter extracton Feature analyss Feature extracton Speaker models Classfyn g Pattern classfcaton o perform PCA we fnd the egenvalues and egenvectors of the sample covarance matrx. Rearrangng the egenvectors n descendng order accordng to the correspondng egenvalues, a lnear transformaton matrx s formed whch generates new vectors from x by ( x µ ) x ' = (6) he egenvectors of C are the prncpal components. In the projected space, the new vectors x ' are mnmally correlated. In order to explot the dmensonalty reducton of PCA we would smply choose the top k (k<m) egenvectors of C to form. hs s the common way of choosng the egenvectors to nclude n the transformaton matrx. An assumpton that s made n PCA dmensonalty reducton s that most of the nformaton contaned n the observaton vectors can be adequately represented n the subspace spanned by the frst k prncpal components. a.) Shortcomngs of PCA PCA s a well-establshed technque n pattern recognton and research nto ts applcaton s worthwhle. III. AUOMAIC SPEAKER RECOGIIO Automatc speaker recognton s task of usng a computer to determne who an utterance was spoken by usng a sample of ther speech as a bometrc measure. Speaker recognton dvdes nto two categores, speaker dentfcaton and speaker verfcaton. In ths paper we are concerned wth the latter whch, as the name suggests, s the process of authentcatng whether a speaker s who they clam to be. he degree of smlarty between the test speech sample and the clamed speaker s model s compared to some predefned threshold and the result used to decde whether to accept or reject the clamant. hs task then also dvdes nto text-dependent and text-ndependent speaker verfcaton. In text-ndependent speaker verfcaton the content of the speech used for the verfcaton s a pror unknown. hs type of authentcaton system s more suted to hghly securty senstve applcatons as the rsk of an mpostor fndng out the requred text and ganng access to the system. Fgure 2 shows a smplfed speaker verfcaton system. he detaled dscusson of all the components of the system s beyond the scope of ths paper. However, the nterested reader s referred to [9,0]. Fgure 2: A smplfed speaker verfcaton system he speaker verfcaton s a two class problem n that we are tryng to determne whether the test speech belongs to the clamed speaker (class +) or to an mpostor set (class -). he mpostor set s generally approxmated by speech data from several possble mpostors. A test utterance s compared to the speakers model and the mpostor set and the classfer then decdes whether the clamant s who they clam to be or an mpostor. a.) Measurng System Performance In speaker verfcaton there are two types of errors that may occur; false acceptance (FA) and false rejecton (FR). An FA error refers to the case when an mpostor s classfed as a authentc system user. Applcatons that requre hgh securty am to keep these errors at a mnmum n order to protect the system from unauthorzed use. he FR error occurs when an authorzed user s wrongly classfed as an mpostor and thus dened access to the system. here s generally a tradeoff between FA and FR errors. here a common way of measurng the performance of a speaker verfcaton system s by defnng the pont where rate of FR s s equal to the rate of FA s. hs pont s known as the Equal Error Rate (ERR). he Detecton Error radeoff (DE) [] curve s a popular way of graphcally representng the expected performance of speaker verfcaton system. he curve s a plot of the rate of the errors mentoned above, on a devate scale. An advantage of the DE curve s that t lends tself to easy nterpretaton, the close the system s DE curve moves to the orgn, the better the system performance. IV. Identty clam EXPERIMEAL SEUP AD RESULS As mentoned, speaker verfcaton s a two class problem. hs suggests that SVM, whch s nherently a two-class classfer, would be a natural choce for ths task. It has already been successfully appled to ths task n [,2,2]. In ths secton present the expermental setup and a dscusson of the results obtaned. a.) Expermental Setup For the purposes of our study we followed very closely the expermental setup used n [] when buldng up our baselne.

4 hat s, we used2 lnear predctve coeffcents (LPC) from whch 8 cepstral coeffcents (LPCC) and ther deltas were computed whch resulted n a 36-dmensonal feature vector. A smple, energy based voce actvty detecton was used to remove those frames wth energy levels below a certan threshold and mean and varance normalzaton was appled. We dd however make use of a RBF kernel functon for the SVM classfer whch dffers from []. he SVM classfer used s part of the orch machne learnng lbrary from the IDIAP Research Insttute. he speaker verfcaton experments were carred out on the IS 2000 database. he mpostor model was created usng the IS 999 database so that none of the testng data was used n the mpostor model. hs way we avod ntroducng any sort of bas to the system. All the experments were conducted on a 3.2GHz Pentum 4 processor. b.) Results he results below compare the performance of the baselne system (whch uses a 36 dmensonal feature vector) to the results acheved when Prncpal Component Analyss feature extracton and dmensonalty reducton was appled. ABLE : Results of feature extracton and dmensonalty reducton on SVM Speaker Verfcaton System Average ranng me [seconds] Average estng me [seconds] Baselne d PCA d PCA d PCA Mss probablty (n %) Result Comparson 36-d PCA Baselne 32-d PCA 24-d PCA False Alarm probablty (n %) Fgure 3: DE plot comparng results V. DISCUSSIO OF RESULS able shows the average tranng and testng tmes n seconds. he baselne system had the longest tranng tme whle the systems whch employed PCA feature extracton showed a sgnfcant decrease n both tranng and testng tmes. Fgure 3 shows the resultant DE curves. he baselne outperforms the applcaton of PCA. A possble reason of ths could be that by applyng the PCA transformaton we change the structure of the data. As mentoned, the choce of kernel and kernel parameters for SVM s data dependent. hus t s lkely that the transformed data could be better classfed by a dfferent kernel functon. VI. COCLUSIOS AD FUURE WORK he results show that applyng feature extracton, even wthout dmensonalty reducton, decreases the tranng and testng tmes. However there s also degradaton n performance when PCA s appled. he savngs n processng tme make t worthwhle to explore the possblty of mprovng the system performance whle applyng feature extracton. A few possbltes for achevng are noted below. he choce of optmal kernel for any partcular SVM task remans a matter of tral and error. here s no formula to determne whch kernel s most approprate for whch task. hus t may be necessary to change the kernel parameters or maybe even the kernel tself n order to obtan optmum performance on data that has been transformed by PCA or any other transformaton algorthm. In ths study we chose to use the tradtonal approach of usng the egenvalues as the crtera for choosng the egenvectors wth whch to buld the transformaton matrx for PCA. In future work we am to repeat the experment, havng optmzed the system parameters, usng the Fsher rato n order to determne whether ths method wll perform as well on an SVM based system as t dd on the VQ based system n [4]. Other future work ncludes nvestgatng the performance of other feature extracton technques such as Independent Component Analyss when appled to ths task. VII. REFERECES [.] W. M. Campbell, J. P. Campbell, D. A. Reynolds, E. Snger and P. A. orres-carrasqullo, Support Vector Machnes for Speaker and Language Recognton. Computer Speech and Language, August [2.] V. Wan, Speaker Verfcaton wth Support Vector Machnes. PhD thess, Department of Computer Scence, Department of Sheffeld, June [3.] L. Wolf and S. Blesch, Combnng Varable Selecton

5 wth Dmensonalty Reducton. Massachusetts Insttute of echnology, Computer Scence and Artfcal Intellgence Laboratory, CBCL Memo 247, March 2005 [4.] P. Dng and L. Zhang, Speaker Recognton usng Prncpal Component Analyss. In Proceedngs of ICOIP 200. Shangha, Chna. ovember 200 [5.] X. Wang, Feature Extracton and Dmensonalty Reducton n Pattern Recognton and her Applcaton n Speech Recognton. PhD thess, School of Mcroelectroncal Engneerng, Grffth Unversty, ovember [6.] M. Awad and L.Khan, Applcatons and Lmtatons of Support Vector Machnes. Department of Computer Scence, Unversty of exas at Dallas, USA. [7.] A. Ganapathraju, Support Vector Machnes for Speech Recognton. PhD thess, Department of Electrcal and Computer Engneerng, Mssssp State Unversty, May [8.] P. Watanachaturaporn, P.K. Varshney and M. K. Arora, Evaluaton of Factors Affectng Support Vector Machnes for Hyperspectral Classfcaton. In Proceedngs of ASPRS Baltmore, USA, March [9.] C. J. C. Burges, A utoral on Support Vector Machnes for Pattern Recognton. Data Mnng and Knowledge Dscovery,998, vol.2, pg [0.]F. Bmbot, J-F. Bonastre, C. Fredoulle, G. Graver, I Magrn-Chagnolleau, S. Megner,. Merln, J. Ortega- Garca, D. Petrovska-Delacrétaz and D. A. Reynolds, A utoral on ext-independent Speaker Verfcaton. Journal on Appled Sgnal Processng 2004 vol 4, pg [.]A. Martn, G. Doddngton,. Kamm, M. Ordowsk and M. Przybock, he DE Curve n Assesment of Detecton ask Performance. [2.]W. M. Campbell, J. P. Campbell, D. A. Reynolds, D. A. Jones and. R. Leek, Phonetc Speaker Recognton wth Support Vector Machnes. MI Lncoln Laboratory, Lexngton. [3.]M.E. Wall, A. Rechtstener and L.M. Rocha, Sngular Value Decomposton and Prncpal Component Analyss. In A Practcal Approach to Mcroarray Data Analyss, (edtors D.P. Berrar, W. Dubtzky and M. Granzow. Kluwer; orwell, MA, 2003, pg [4.]R. Collobert and S. Bengo, SVMorch: Support Vector Machnes for Large-Scale Regresson Problems. Journal of Machne Learnng Research, 200, vol., pg

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

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

Face Recognition Based on SVM and 2DPCA

Face Recognition Based on SVM and 2DPCA Vol. 4, o. 3, September, 2011 Face Recognton Based on SVM and 2DPCA Tha Hoang Le, Len Bu Faculty of Informaton Technology, HCMC Unversty of Scence Faculty of Informaton Scences and Engneerng, Unversty

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

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

Machine Learning. Support Vector Machines. (contains material adapted from talks by Constantin F. Aliferis & Ioannis Tsamardinos, and Martin Law)

Machine Learning. Support Vector Machines. (contains material adapted from talks by Constantin F. Aliferis & Ioannis Tsamardinos, and Martin Law) Machne Learnng Support Vector Machnes (contans materal adapted from talks by Constantn F. Alfers & Ioanns Tsamardnos, and Martn Law) Bryan Pardo, Machne Learnng: EECS 349 Fall 2014 Support Vector Machnes

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

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

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

Announcements. Supervised Learning

Announcements. Supervised Learning Announcements See Chapter 5 of Duda, Hart, and Stork. Tutoral by Burge lnked to on web page. Supervsed Learnng Classfcaton wth labeled eamples. Images vectors n hgh-d space. Supervsed Learnng Labeled eamples

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

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

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1

Outline. Discriminative classifiers for image recognition. Where in the World? A nearest neighbor recognition example 4/14/2011. CS 376 Lecture 22 1 4/14/011 Outlne Dscrmnatve classfers for mage recognton Wednesday, Aprl 13 Krsten Grauman UT-Austn Last tme: wndow-based generc obect detecton basc ppelne face detecton wth boostng as case study Today:

More information

Classification of Face Images Based on Gender using Dimensionality Reduction Techniques and SVM

Classification of Face Images Based on Gender using Dimensionality Reduction Techniques and SVM Classfcaton of Face Images Based on Gender usng Dmensonalty Reducton Technques and SVM Fahm Mannan 260 266 294 School of Computer Scence McGll Unversty Abstract Ths report presents gender classfcaton based

More information

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification

12/2/2009. Announcements. Parametric / Non-parametric. Case-Based Reasoning. Nearest-Neighbor on Images. Nearest-Neighbor Classification Introducton to Artfcal Intellgence V22.0472-001 Fall 2009 Lecture 24: Nearest-Neghbors & Support Vector Machnes Rob Fergus Dept of Computer Scence, Courant Insttute, NYU Sldes from Danel Yeung, John DeNero

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

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

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION

MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION MULTISPECTRAL IMAGES CLASSIFICATION BASED ON KLT AND ATR AUTOMATIC TARGET RECOGNITION Paulo Quntlano 1 & Antono Santa-Rosa 1 Federal Polce Department, Brasla, Brazl. E-mals: quntlano.pqs@dpf.gov.br and

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

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

Using Neural Networks and Support Vector Machines in Data Mining

Using Neural Networks and Support Vector Machines in Data Mining Usng eural etworks and Support Vector Machnes n Data Mnng RICHARD A. WASIOWSKI Computer Scence Department Calforna State Unversty Domnguez Hlls Carson, CA 90747 USA Abstract: - Multvarate data analyss

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

Feature Extraction and Test Algorithm for Speaker Verification

Feature Extraction and Test Algorithm for Speaker Verification Feature Extracton and Test Algorthm for Speaker Verfcaton Wu Guo, Renhua Wang and Lrong Da Unversty of Scence and Technology of Chna, Hefe guowu@mal.ustc.edu.cn,{rhw, lrda}@ustc,edu.cn Abstract. In ths

More information

Machine Learning 9. week

Machine Learning 9. week Machne Learnng 9. week Mappng Concept Radal Bass Functons (RBF) RBF Networks 1 Mappng It s probably the best scenaro for the classfcaton of two dataset s to separate them lnearly. As you see n the below

More information

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach

Skew Angle Estimation and Correction of Hand Written, Textual and Large areas of Non-Textual Document Images: A Novel Approach Angle Estmaton and Correcton of Hand Wrtten, Textual and Large areas of Non-Textual Document Images: A Novel Approach D.R.Ramesh Babu Pyush M Kumat Mahesh D Dhannawat PES Insttute of Technology Research

More information

Human Face Recognition Using Generalized. Kernel Fisher Discriminant

Human Face Recognition Using Generalized. Kernel Fisher Discriminant Human Face Recognton Usng Generalzed Kernel Fsher Dscrmnant ng-yu Sun,2 De-Shuang Huang Ln Guo. Insttute of Intellgent Machnes, Chnese Academy of Scences, P.O.ox 30, Hefe, Anhu, Chna. 2. Department of

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

BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET

BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET 1 BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET TZU-CHENG CHUANG School of Electrcal and Computer Engneerng, Purdue Unversty, West Lafayette, Indana 47907 SAUL B. GELFAND School

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

CLASSIFICATION OF ULTRASONIC SIGNALS

CLASSIFICATION OF ULTRASONIC SIGNALS The 8 th Internatonal Conference of the Slovenan Socety for Non-Destructve Testng»Applcaton of Contemporary Non-Destructve Testng n Engneerng«September -3, 5, Portorož, Slovena, pp. 7-33 CLASSIFICATION

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

Detection of an Object by using Principal Component Analysis

Detection of an Object by using Principal Component Analysis Detecton of an Object by usng Prncpal Component Analyss 1. G. Nagaven, 2. Dr. T. Sreenvasulu Reddy 1. M.Tech, Department of EEE, SVUCE, Trupath, Inda. 2. Assoc. Professor, Department of ECE, SVUCE, Trupath,

More information

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task

Term Weighting Classification System Using the Chi-square Statistic for the Classification Subtask at NTCIR-6 Patent Retrieval Task Proceedngs of NTCIR-6 Workshop Meetng, May 15-18, 2007, Tokyo, Japan Term Weghtng Classfcaton System Usng the Ch-square Statstc for the Classfcaton Subtask at NTCIR-6 Patent Retreval Task Kotaro Hashmoto

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

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

A Modified Median Filter for the Removal of Impulse Noise Based on the Support Vector Machines

A Modified Median Filter for the Removal of Impulse Noise Based on the Support Vector Machines A Modfed Medan Flter for the Removal of Impulse Nose Based on the Support Vector Machnes H. GOMEZ-MORENO, S. MALDONADO-BASCON, F. LOPEZ-FERRERAS, M. UTRILLA- MANSO AND P. GIL-JIMENEZ Departamento de Teoría

More information

Network Intrusion Detection Based on PSO-SVM

Network Intrusion Detection Based on PSO-SVM TELKOMNIKA Indonesan Journal of Electrcal Engneerng Vol.1, No., February 014, pp. 150 ~ 1508 DOI: http://dx.do.org/10.11591/telkomnka.v1.386 150 Network Intruson Detecton Based on PSO-SVM Changsheng Xang*

More information

Face Recognition Method Based on Within-class Clustering SVM

Face Recognition Method Based on Within-class Clustering SVM Face Recognton Method Based on Wthn-class Clusterng SVM Yan Wu, Xao Yao and Yng Xa Department of Computer Scence and Engneerng Tong Unversty Shangha, Chna Abstract - A face recognton method based on Wthn-class

More information

RECOGNIZING GENDER THROUGH FACIAL IMAGE USING SUPPORT VECTOR MACHINE

RECOGNIZING GENDER THROUGH FACIAL IMAGE USING SUPPORT VECTOR MACHINE Journal of Theoretcal and Appled Informaton Technology 30 th June 06. Vol.88. No.3 005-06 JATIT & LLS. All rghts reserved. ISSN: 99-8645 www.jatt.org E-ISSN: 87-395 RECOGNIZING GENDER THROUGH FACIAL IMAGE

More information

Modular PCA Face Recognition Based on Weighted Average

Modular PCA Face Recognition Based on Weighted Average odern Appled Scence odular PCA Face Recognton Based on Weghted Average Chengmao Han (Correspondng author) Department of athematcs, Lny Normal Unversty Lny 76005, Chna E-mal: hanchengmao@163.com Abstract

More information

Incremental Learning with Support Vector Machines and Fuzzy Set Theory

Incremental Learning with Support Vector Machines and Fuzzy Set Theory The 25th Workshop on Combnatoral Mathematcs and Computaton Theory Incremental Learnng wth Support Vector Machnes and Fuzzy Set Theory Yu-Mng Chuang 1 and Cha-Hwa Ln 2* 1 Department of Computer Scence and

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

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

Feature Extractions for Iris Recognition

Feature Extractions for Iris Recognition Feature Extractons for Irs Recognton Jnwook Go, Jan Jang, Yllbyung Lee, and Chulhee Lee Department of Electrcal and Electronc Engneerng, Yonse Unversty 134 Shnchon-Dong, Seodaemoon-Gu, Seoul, KOREA Emal:

More information

An Anti-Noise Text Categorization Method based on Support Vector Machines *

An Anti-Noise Text Categorization Method based on Support Vector Machines * An Ant-Nose Text ategorzaton Method based on Support Vector Machnes * hen Ln, Huang Je and Gong Zheng-Hu School of omputer Scence, Natonal Unversty of Defense Technology, hangsha, 410073, hna chenln@nudt.edu.cn,

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

A Statistical Model Selection Strategy Applied to Neural Networks

A Statistical Model Selection Strategy Applied to Neural Networks A Statstcal Model Selecton Strategy Appled to Neural Networks Joaquín Pzarro Elsa Guerrero Pedro L. Galndo joaqun.pzarro@uca.es elsa.guerrero@uca.es pedro.galndo@uca.es Dpto Lenguajes y Sstemas Informátcos

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

Online Detection and Classification of Moving Objects Using Progressively Improving Detectors

Online Detection and Classification of Moving Objects Using Progressively Improving Detectors Onlne Detecton and Classfcaton of Movng Objects Usng Progressvely Improvng Detectors Omar Javed Saad Al Mubarak Shah Computer Vson Lab School of Computer Scence Unversty of Central Florda Orlando, FL 32816

More information

Discriminative Projection Selection Based Face Image Hashing

Discriminative Projection Selection Based Face Image Hashing IEICE TRANS. INF. & SYST., VOL.Exx D, NO.xx XXXX 00x LETTER Dscrmnatve Projecton Selecton Based Face Image Hashng Cagatay KARABAT a), Student Member and Hakan ERDOGAN, Nonmember SUMMARY Face mage hashng

More information

BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION

BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION SHI-LIANG SUN, HONG-LEI SHI Department of Computer Scence and Technology, East Chna Normal Unversty 500 Dongchuan Road, Shangha 200241, P. R. Chna E-MAIL: slsun@cs.ecnu.edu.cn,

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

A Robust LS-SVM Regression

A Robust LS-SVM Regression PROCEEDIGS OF WORLD ACADEMY OF SCIECE, EGIEERIG AD ECHOLOGY VOLUME 7 AUGUS 5 ISS 37- A Robust LS-SVM Regresson József Valyon, and Gábor Horváth Abstract In comparson to the orgnal SVM, whch nvolves a quadratc

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

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like:

Outline. Self-Organizing Maps (SOM) US Hebbian Learning, Cntd. The learning rule is Hebbian like: Self-Organzng Maps (SOM) Turgay İBRİKÇİ, PhD. Outlne Introducton Structures of SOM SOM Archtecture Neghborhoods SOM Algorthm Examples Summary 1 2 Unsupervsed Hebban Learnng US Hebban Learnng, Cntd 3 A

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

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

Feature Selection as an Improving Step for Decision Tree Construction

Feature Selection as an Improving Step for Decision Tree Construction 2009 Internatonal Conference on Machne Learnng and Computng IPCSIT vol.3 (2011) (2011) IACSIT Press, Sngapore Feature Selecton as an Improvng Step for Decson Tree Constructon Mahd Esmael 1, Fazekas Gabor

More information

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition

Novel Pattern-based Fingerprint Recognition Technique Using 2D Wavelet Decomposition Mathematcal Methods for Informaton Scence and Economcs Novel Pattern-based Fngerprnt Recognton Technque Usng D Wavelet Decomposton TUDOR BARBU Insttute of Computer Scence of the Romanan Academy T. Codrescu,,

More information

Kent State University CS 4/ Design and Analysis of Algorithms. Dept. of Math & Computer Science LECT-16. Dynamic Programming

Kent State University CS 4/ Design and Analysis of Algorithms. Dept. of Math & Computer Science LECT-16. Dynamic Programming CS 4/560 Desgn and Analyss of Algorthms Kent State Unversty Dept. of Math & Computer Scence LECT-6 Dynamc Programmng 2 Dynamc Programmng Dynamc Programmng, lke the dvde-and-conquer method, solves problems

More information

Feature Selection By KDDA For SVM-Based MultiView Face Recognition

Feature Selection By KDDA For SVM-Based MultiView Face Recognition SEI 007 4 rth Internatonal Conference: Scences of Electronc, echnologes of Informaton and elecommuncatons March 5-9, 007 UNISIA Feature Selecton By KDDA For SVM-Based MultVew Face ecognton Seyyed Majd

More information

PCA Based Gait Segmentation

PCA Based Gait Segmentation Honggu L, Cupng Sh & Xngguo L PCA Based Gat Segmentaton PCA Based Gat Segmentaton Honggu L, Cupng Sh, and Xngguo L 2 Electronc Department, Physcs College, Yangzhou Unversty, 225002 Yangzhou, Chna 2 Department

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

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

EYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS

EYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS P.G. Demdov Yaroslavl State Unversty Anatoly Ntn, Vladmr Khryashchev, Olga Stepanova, Igor Kostern EYE CENTER LOCALIZATION ON A FACIAL IMAGE BASED ON MULTI-BLOCK LOCAL BINARY PATTERNS Yaroslavl, 2015 Eye

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

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

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

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces

Range images. Range image registration. Examples of sampling patterns. Range images and range surfaces Range mages For many structured lght scanners, the range data forms a hghly regular pattern known as a range mage. he samplng pattern s determned by the specfc scanner. Range mage regstraton 1 Examples

More information

Optimizing Document Scoring for Query Retrieval

Optimizing Document Scoring for Query Retrieval Optmzng Document Scorng for Query Retreval Brent Ellwen baellwe@cs.stanford.edu Abstract The goal of ths project was to automate the process of tunng a document query engne. Specfcally, I used machne learnng

More information

Facial Expression Recognition Based on Local Binary Patterns and Local Fisher Discriminant Analysis

Facial Expression Recognition Based on Local Binary Patterns and Local Fisher Discriminant Analysis WSEAS RANSACIONS on SIGNAL PROCESSING Shqng Zhang, Xaomng Zhao, Bcheng Le Facal Expresson Recognton Based on Local Bnary Patterns and Local Fsher Dscrmnant Analyss SHIQING ZHANG, XIAOMING ZHAO, BICHENG

More information

Discriminative classifiers for object classification. Last time

Discriminative classifiers for object classification. Last time Dscrmnatve classfers for object classfcaton Thursday, Nov 12 Krsten Grauman UT Austn Last tme Supervsed classfcaton Loss and rsk, kbayes rule Skn color detecton example Sldng ndo detecton Classfers, boostng

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

PERFORMANCE EVALUATION FOR SCENE MATCHING ALGORITHMS BY SVM

PERFORMANCE EVALUATION FOR SCENE MATCHING ALGORITHMS BY SVM PERFORMACE EVALUAIO FOR SCEE MACHIG ALGORIHMS BY SVM Zhaohu Yang a, b, *, Yngyng Chen a, Shaomng Zhang a a he Research Center of Remote Sensng and Geomatc, ongj Unversty, Shangha 200092, Chna - yzhac@63.com

More information

Wavelets and Support Vector Machines for Texture Classification

Wavelets and Support Vector Machines for Texture Classification Wavelets and Support Vector Machnes for Texture Classfcaton Kashf Mahmood Rapoot Faculty of Computer Scence & Engneerng, Ghulam Ishaq Khan Insttute, Top, PAKISTAN. kmr@gk.edu.pk Nasr Mahmood Rapoot Department

More information

A Background Subtraction for a Vision-based User Interface *

A Background Subtraction for a Vision-based User Interface * A Background Subtracton for a Vson-based User Interface * Dongpyo Hong and Woontack Woo KJIST U-VR Lab. {dhon wwoo}@kjst.ac.kr Abstract In ths paper, we propose a robust and effcent background subtracton

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

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

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

A Deflected Grid-based Algorithm for Clustering Analysis

A Deflected Grid-based Algorithm for Clustering Analysis A Deflected Grd-based Algorthm for Clusterng Analyss NANCY P. LIN, CHUNG-I CHANG, HAO-EN CHUEH, HUNG-JEN CHEN, WEI-HUA HAO Department of Computer Scence and Informaton Engneerng Tamkang Unversty 5 Yng-chuan

More information

Meta-heuristics for Multidimensional Knapsack Problems

Meta-heuristics for Multidimensional Knapsack Problems 2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,

More information

An Improved Image Segmentation Algorithm Based on the Otsu Method

An Improved Image Segmentation Algorithm Based on the Otsu Method 3th ACIS Internatonal Conference on Software Engneerng, Artfcal Intellgence, Networkng arallel/dstrbuted Computng An Improved Image Segmentaton Algorthm Based on the Otsu Method Mengxng Huang, enjao Yu,

More information

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System Fuzzy Modelng of the Complexty vs. Accuracy Trade-off n a Sequental Two-Stage Mult-Classfer System MARK LAST 1 Department of Informaton Systems Engneerng Ben-Guron Unversty of the Negev Beer-Sheva 84105

More information

An Image Fusion Approach Based on Segmentation Region

An Image Fusion Approach Based on Segmentation Region Rong Wang, L-Qun Gao, Shu Yang, Yu-Hua Cha, and Yan-Chun Lu An Image Fuson Approach Based On Segmentaton Regon An Image Fuson Approach Based on Segmentaton Regon Rong Wang, L-Qun Gao, Shu Yang 3, Yu-Hua

More information

Comparison Study of Textural Descriptors for Training Neural Network Classifiers

Comparison Study of Textural Descriptors for Training Neural Network Classifiers Comparson Study of Textural Descrptors for Tranng Neural Network Classfers G.D. MAGOULAS (1) S.A. KARKANIS (1) D.A. KARRAS () and M.N. VRAHATIS (3) (1) Department of Informatcs Unversty of Athens GR-157.84

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

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

Data Mining: Model Evaluation

Data Mining: Model Evaluation Data Mnng: Model Evaluaton Aprl 16, 2013 1 Issues: Evaluatng Classfcaton Methods Accurac classfer accurac: predctng class label predctor accurac: guessng value of predcted attrbutes Speed tme to construct

More information

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE Dorna Purcaru Faculty of Automaton, Computers and Electroncs Unersty of Craoa 13 Al. I. Cuza Street, Craoa RO-1100 ROMANIA E-mal: dpurcaru@electroncs.uc.ro

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

Fingerprint matching based on weighting method and SVM

Fingerprint matching based on weighting method and SVM Fngerprnt matchng based on weghtng method and SVM Ja Ja, Lanhong Ca, Pnyan Lu, Xuhu Lu Key Laboratory of Pervasve Computng (Tsnghua Unversty), Mnstry of Educaton Bejng 100084, P.R.Chna {jaja}@mals.tsnghua.edu.cn

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

Improving Classifier Fusion Using Particle Swarm Optimization

Improving Classifier Fusion Using Particle Swarm Optimization Proceedngs of the 7 IEEE Symposum on Computatonal Intellgence n Multcrtera Decson Makng (MCDM 7) Improvng Classfer Fuson Usng Partcle Swarm Optmzaton Kalyan Veeramachanen Dept. of EECS Syracuse Unversty

More information

A Novel Term_Class Relevance Measure for Text Categorization

A Novel Term_Class Relevance Measure for Text Categorization A Novel Term_Class Relevance Measure for Text Categorzaton D S Guru, Mahamad Suhl Department of Studes n Computer Scence, Unversty of Mysore, Mysore, Inda Abstract: In ths paper, we ntroduce a new measure

More information

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning Computer Anmaton and Vsualsaton Lecture 4. Rggng / Sknnng Taku Komura Overvew Sknnng / Rggng Background knowledge Lnear Blendng How to decde weghts? Example-based Method Anatomcal models Sknnng Assume

More information

Backpropagation: In Search of Performance Parameters

Backpropagation: In Search of Performance Parameters Bacpropagaton: In Search of Performance Parameters ANIL KUMAR ENUMULAPALLY, LINGGUO BU, and KHOSROW KAIKHAH, Ph.D. Computer Scence Department Texas State Unversty-San Marcos San Marcos, TX-78666 USA ae049@txstate.edu,

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

Collaboratively Regularized Nearest Points for Set Based Recognition

Collaboratively Regularized Nearest Points for Set Based Recognition Academc Center for Computng and Meda Studes, Kyoto Unversty Collaboratvely Regularzed Nearest Ponts for Set Based Recognton Yang Wu, Mchhko Mnoh, Masayuk Mukunok Kyoto Unversty 9/1/013 BMVC 013 @ Brstol,

More information

Parallel matrix-vector multiplication

Parallel matrix-vector multiplication Appendx A Parallel matrx-vector multplcaton The reduced transton matrx of the three-dmensonal cage model for gel electrophoress, descrbed n secton 3.2, becomes excessvely large for polymer lengths more

More information

Spam Filtering Based on Support Vector Machines with Taguchi Method for Parameter Selection

Spam Filtering Based on Support Vector Machines with Taguchi Method for Parameter Selection E-mal Spam Flterng Based on Support Vector Machnes wth Taguch Method for Parameter Selecton We-Chh Hsu, Tsan-Yng Yu E-mal Spam Flterng Based on Support Vector Machnes wth Taguch Method for Parameter Selecton

More information