The Discriminate Analysis and Dimension Reduction Methods of High Dimension

Size: px
Start display at page:

Download "The Discriminate Analysis and Dimension Reduction Methods of High Dimension"

Transcription

1 Open Journal of Socal Scences, 015, 3, 7-13 Publshed Onlne March 015 n ScRes. The Dscrmnate Analyss and Dmenson Reducton Methods of Hgh Dmenson Lan Fu 1, 1 Statstcs School, Jangx Unversty of Fnance of Economcs, Nanchang, Chna Research Center of Appled Statstcs, Jangx Unversty of Fnance of Economcs, Nanchang, Chna Emal: fulan @sna.com Receved September 014 Abstract Statstcal methods have been gettng constant development snce 1970s. However, the statstcal methods of the bg data are no longer restrcted wth these methods whch are lsted n the textbook. Ths paper manly demonstrates the Dscrmnaton Analyss of Multvarate Statstcal Analyss, Lnear Dmensonalty Reducton and Nonlnear Dmensonalty Reducton Method under the crcumstances of the wde range of applcatons of hgh-dmensonal data. Ths paper ncludes three parts. To begn wth, the paper llustrates a developng trend from the data to the hgh-dmensonal. Meanwhle, t analyzes the mpacts of the hgh-dmensonal data on dscrmnate analyss methods. The second part represents the necessty of the dmensonalty reducton studes n the era of the hgh-dmensonal data overflowng. Then, the paper focuses on ntroducng the man methods of the lnear dmensonalty reducton. In addton, ths paper covers the basc dea of the nonlnear dmensonalty reducton. Moreover, t systematcally analyzes the breakthrough of the tradtonal methods. Furthermore, t chronologcally demonstrates the developng trend of the dmensonalty reducton method. The fnal part shows a comprehensve and systematc concluson to the whole essay and descrbes a developng prospect of the dmenson reducton methods n the future. The purpose of ths essay s to desgn a framework of a performance system whch s subject to the characterstcs of Chna Hgh-tech enterprses. It based on the analysng the prncples and sgnfcance of the performance system of Hgh-tech enterprses. The framework wll promote the standardze management of Hgh-tech enterprses of Chna. Keywords Hgh-Dmensonal Data, Method of Dmensonalty Reducton, Dscrmnant Analyss, Lnear Dmensonalty Reducton, Manfold Structure 1. Introducton The comng century s surely the century of data [1]. People dd have access to a large amount of fnancal data, mage data and space data recently. These data have a common feature whch s they are all hgh-dmensonal data. Actually, most data people obtaned are hgh-dmensonal n the world. The ncrease of data dmenson brngs the Gospel of the dmenson meanwhle leads to the data dsaster. Ths nformaton whch s hdden How to cte ths paper: Fu, L. (015) The Dscrmnate Analyss and Dmenson Reducton Methods of Hgh Dmenson. Open Journal of Socal Scences, 3,

2 by the hgh-dmensonal data provdes a possblty for solvng the scentfc problems. However, t s dffcult to magne the challenges of the data processng technology because of the ncrease of the data dmenson. The method of dmensonalty reducton s defned as a low-dmensonal expresson way whch can fathfully reflect the nherent feature of the raw data n hgh-dmensonal data. It caused the wdespread attenton of people because the dmenson reducton method s consdered as an effectve measure to prevent the curses of dmensonalty [] happenng. So far people had found a lot of dfferent dmenson reducton methods whch nvolve the Projecton Pursut Method, Prncple Component Analyss (PCA) and Manfold Learnng. However, the Isomer, Locally Lnear Embeddng (LLE) and Palladan Egen map are all based on the Manfold Learnng [3]. There are several dfferent methods of dvson for the dmenson reducton methods. Accordng to the nature of data, the method of dmenson reducton can be dvded nto Lnear Dmensonalty Reducton and Nonlnear Dmensonalty Reducton. In terms of the geometrcal structure of the preservaton degree of the dmenson reducton, t ncludes local approach and global approach. These dmenson reducton methods obtan systematcal analyss and contrast.. Hgh-Dmensonal Tendency of the Data and Its Impact.1. The Dervatve of Hgh-Dmensonal Data Wth the development of the scence and technology, people can obtan a great deal of nformaton to reveal the objectve law whch s covered by superfcal phenomenon. Ths leads to a tendency whch s the data s gradually development from the low-dmenson to the hgh-dmenson. In other words, the mprovement of data dmenson brngs the Gospel of the dmenson meanwhle result n the data dsaster [4]. The curses of dmensonalty was frst put forward by Bellman n At that tme, the curses of dmensonalty refers to the ncensement of the dmenson would lead to an exponental growth of samples n certan statstcal ndex model. Meanwhle, the ndcatng length and complexty of the processng nformaton would have a exponental growth. Nowadays the curses of dmensonalty manly refer to the ntrnsc sparsely n the hgh dmensonal data space. In other words, t s the spatal representaton n the hgh dmensonal data space. The curses of dmensonalty s caused by the ntrnsc dmenson rather than the representaton dmenson. It affects some felds whch nvolve Statstcal Estmaton, Numercal Integraton, Optmzaton Problems, and Probablty Densty Estmaton so on. Thus, the dmenson reducton method as a low-dmensonal expresson method of the hgh-dmensonal data gettng the essental characterstcs of the orgnal date can effectvely avod the curses of dmensonalty. It causes the wdespread attenton of people... Impact of Hgh Dmensonalty for Classfcaton There s a common problem n the process of classfcaton. The sze of the dmenson p of feature vectors s much greater than the sze of the tranng sample n. In addton, only a part of p feature vectors plays a mportant role n the classfcaton. Although the tradtonal analyss method s extremely mportant, t s not sutable for hgh-dmensonal case. The followng statement llustrates the nfluences of the hgh-dmenson for the Fsher dscrmnate analyss and ndependent classfcaton crtera respectvely. Besdes, the statement assumes the dscrmnaton occurs n the two hgh-dmensonal populatons. And the probablty of the two populatons s 50%. Meanwhle, there s comparablty between Sample n 1 and smple n...1. Fsher Dscrmnaton Based on the Hgh Dmensonalty 1 The Fsher dscrmnate s δ ( x ) ( x u T F ) ( u u 1 = Σ 1 ). It can assure both Σ and Σ are non ll-condtoned matrx n the parameter space. It also can guarantee one of the Mahalanobs dstances of two categores s postve C. Therefore, t can the utlzes ˆF δ of the sample data estmaton to obtan the asymptotc optmalty n low-dmenson stuaton. However, the stuaton s dfferent from hgh-dmenson. One reason s that the true covarance matrx s not ll-condtoned matrx. Another reason s that the abnormal value of the sample covarance s not sutable for the Fsher crteron because the stuaton of the sze of dmensonalty s greater than the sze of the sample.... Impact of the Dmensonalty for the Independent Crtera The dffcultes of the hgh-dmensonal dscrmnated le n a lot of nose characterstcs. These nose characte- 8

3 rstcs have no any functon for the reducton of the classfcaton errors. There s another reason for the poor results of the Fsher dscrmnate under the hgh-dmensonalty crcumstances. That s people can ndvdually estmates for each parameter. However, the error wll be great as people estmate multple feature estmaton at the same tme. Therefore, ths wll lead to the sgnfcant ncrease of the error probablty [5]. 3. The Man Method of the Dmenson Reducton 3.1. The Classfcaton of the Dmenson Reducton Method In the lght of the nature of the data, the method of dmenson reducton can be dvded nto Lnear Dmensonalty Reducton and Nonlnear Dmensonalty Reducton. These two methods can be further dvded nto some manstreamng classfcaton methods and the detals [6]. As shown Fgure The Man Methods of the Lnear Dmenson Reducton The lnear dmenson reducton method s most wdely applcaton n the dmensonalty reducton. There are some excellent propertes n the lnear dmenson reducton method. These propertes nvolve easly explcable, an analytcal solutons, smple calculaton and the effectveness of the lnear structure of the data collecton. As matter of fact, the lnear dmensonalty reducton method s based on the dfferent optmzaton crteron. It seeks the best lnear model. Ths s totally dfferent wth nonlnear dmensonalty reducton method. However, t s the common property of all lnear dmensonalty reducton methods. The lnear dmensonalty reducton methods manly nclude the Prncpal Component Analyss and the Lnear Dscrmnate Analyss. The nonlnear dmensonalty reducton methods [7]-[9] chefly nvolve the Locally Lnear Embeddng, Palladan Egen maps, Isomer, Multdmensonal Scale Method and the Local Tangent Space Method so on Prncpal Component Analyss (PCA) The Prncpal Component Analyss [10] was frst put forward by Pearson (K.Pearson) n However, Hostellng (H. Hostellng) establshed a mathematcal foundaton for PCA n The theoretcal bass of PCA s based on the karhunen-levy expanson. The basc dea of PCA s that the orgnal varables can be transformed nto several comprehensve varables whch are able to reflect the man propertes of the thngs. Meanwhle, the hgh-dmensonal data can be shadowed on the low-dmensonal space. Ths ensures a most effectve low-dmensonal data n the least square sense and truly represents the orgnal hgh-dmensonal data. Ths method focuses on a fdelty optmzaton of the orgnal hgh-dmensonal data. It tres to solve a problem of the projecton drecton and obtan the representaton of the orgnal data n the least square sense. From the mathematcal pont of vew, the startng pont of the prncpal component analyss s the sample covarance matrx S. Fgure 1. The classfcaton of the dmenson reducton method. 9

4 s1 s1... s 1p s1 s... s p S = s1p sp sp s s the varance of the varable x. S j s the covarance between varable x and varable x j. There s no lnear correlaton between the varables x and the varables x j f the covarance s zero. The strength of the lnear correlaton between the varables x and the varables x j can be expressed by the correlaton coeffcent R y whch s equal sj ( ss j ). The change of the man spndle would transform the number of p correlatve varables ( x1, x, xp ) nto the number of p uncreatve varables ( z1, z, zp ). The varable of the new axes descrbed by an egenvector u. T z = u ( x X) That s the th prncpal components. (The mean of Z s zero, the varance s l whch s the th egenvalue.) There are some man steps of the dmenson reducton. The frst step s to obtan the egenvalues and the egenvectors of the covarance matrx, then rank the sze of the egenvalues. The second step s to choose the egenvector of the bg egenvalues correspondng and put ths egenvector as a projecton vector. The fnal step s the hgh-dmensonal data would be projected onto the subspace of the projecton vector spanned. Undoubtfully, there are some problems and lmtatons on the PCA. Frst of all, the PCA have to assume the approxmate normalty of the nput space. Secondly, the PCA have to assume that the nput data s real and contnuous. Thrdly, the PCA wll fal f the dstrbuton of the data had the complex manfold structure Lnear Dscrmnaton Analyss (LDA) The Lnear Dscrmnaton Analyss was frst put forward by Fsher (R.Fsher) n The LDA s smlar to the PCA. However, there are some dfferences between the LDA and the PCA. There are some common ponts between LDA and LDA. Frstly, LDA and PCA both are obtaned by a set of projecton vectors, and utlze the hgh-dmenson data to project the low-dmensonal space. Therefore, LDA and PCA have the smlar procedures. The second pont s that the same purpose of LDA and PCA both are dmensonalty reducton. Obvously, there are some dfferences between LDA and PCA. Frst of all, the projecton vectors of LDA and PCA are totally dfferent. The reason s that the key pont of LDA and PCA are dfferent n the process of dmensonalty reducton. PCA make the optmzaton of the orgnal hgh-dmenson data fdelty of the dmensonalty reducton data as the key. It tres to get a drecton of an optmal projecton. Then the projecton data can represent the orgnal data as far as possble n the constrant condtons. However, LDA focus on the optmzaton of the dfferent data dscrmnaton of the dmensonalty reducton data. The purpose of the LDA s Tyr as far as possble to separate the two types of data. 3.3.The Major Methods of the Nonlnear Dmenson Reducton Local Lnear Embeddng (LLE) The man steps of the Local Lnear Embeddng [11] [1] (LLE) nclude the followng ponts. The frst step s to fnd a set of the K-nearest neghbors of each pont. It means to compute a set of weghts for each pont that best descrbe the pont as a lnear combnaton of ts neghbors. Then t uses an egenvector-based optmzaton technque to fnd the low-dmensonal embeddng of ponts, such that each pont s stll descrbed wth the same lnear combnaton of ts neghbors. The Eucldean dstance and Djkstra dstance of keepng the surface characterstcs of sample ponts can be used n the dstance computaton. The second step s to get a local reconstructon weght matrx of the sample through computng the neghbor pont of each sample. Then t constructs a cost functon to measure the reconstructon error. The cost functon s expressed as ε ( W) = X- XW j j j 10

5 The W of the reconstructon weght s mnmzed under two constrants. One of the constrants s the W wll be zero f pont x s not a neghbor of the pont x j. Another constrant s W j=1. In other words, t s to get W. It means to solve the problem of the least square wth constrant. j The thrd step s to compute the output value of the sample by the local reconstructon weght matrx and the neghbor pont of the sample. Each sample x wll be mapped onto the low-dmensonal vector Y. The man deas of LLE nclude the followng aspects. The frst aspect s to assume each sample pont and ts neghbor pont n a lnear regon of manfold. Another one s to transform the global nonlnear nto the local lnear and the local overlap felds of each sample and neghborng pont can supply the global structure nformaton Local Tangent Space Algnment (LTSA) The emergence of the LTSA (014) s later than other dmensonalty reducton methods. The bass of the LTSA algorthm whch s proposed by Zhang Zhenyue and other researchers nclude the followng parts. One bass s that the LTSA can obtan a new global nonlnear structure by ntegratng the local lnear nformaton and local lnear analyss. Ths structure has the propertes of a nonlnear manfold. The LTSA ncludes two steps whch are projecton and ntegraton. The LTSA can properly show the good propertes of the geometry feature of the manfold. But t also has lmtatons. These lmtatons ndcate that the LTSA cannot do very well to the manfold learnng for hgh curvature Dmensonalty Reducton From the Begnnng to Nowadays The study of the dmensonalty reducton can be traced to The man methods of the dmensonalty reducton nclude Lnear Dscrmnate Analyss, Prncpal Component Analyss and Projecton Pursut so on. The developed dmensonalty reducton methods n recent years manly nvolve the Nonlnear Dmensonalty reducton (It s based on kernel.), the dmensonalty reducton of the two-dmensonalty and tensor, the Manfold Learnng and localzed dmensonalty reducton as well as Sem-supervsed Dmensonalty Reducton Nonlnear Dmensonalty Reducton Based on the Kernel Snce Schölkopf (B. Schölkopf) and other researchers ntroduced the kernel methods to the feld of the dmensonalty reducton, they proposed the classc Kernel Prncpal Component Analyss. In addton, researchers combned wth the kernel methods and other tradtonal dmensonalty reducton methods. Then they proposed some dscrmnate analyss methods. These methods nclude Fsher Dscrmnate Analyss, Kernel Canoncal Correlaton Analyss and Kernel Independent Component Analyss. So far, almost all of the lnear dmensonalty reducton methods have correspondng kemelzed verson. The kernel methods have become standard practce methods for a lnear dmensonalty reducton transformng nto a no lnearzaton dmensonalty reducton. The key pont s the selecton of the kernel and the kernel parameter n the dmenson reducton algorthm whch s based on the kernel. Researchers also focus on mprovng the operatonal effcency of the kernel dmensonalty reducton algorthm n the bg data Two-Dmensonalzaton and Tensor Dmensonalty Reducton The tradtonal model s defned as a sample pont whch s represented as a vector model of a pont n n-dmensonal space. It s obvously that the tradtonal model s more easly to be handled by varous statstcal methods. However, the tradtonal model s no longer sutable for applcaton n practcal under ths stuaton. Ths stuaton s the sample s usually non-vector model and wll transform two dmensonal models nto correspondng vector models. Ths stuaton would lead to the damage of the structure nformaton of the orgnal matrx and the adverse consequences of ncreased tme cost and storage cost. Therefore, the concept of the tensor dmensonalty reducton was proposed. The dea of two-dmensonal matrx mage feature extracton was frst put forward by Lu (K. Lu). Ths dea avods the above adverse consequences. That s people can drectly wthdraw the egenvalues and not necessary to pull a mage nto a vector. Snce then, the method had access to the development and research from dfferent angles. The Two-Dmensonal Prncpal Component Analyss (DPCA), Two-Dmensonal Lnear Dscrmnate Analyss (DLDA) and Tow-dmensonal Canoncal Correlaton Analyss (DCCA) were proposed. The methods of two dmensonal- 11

6 zaton and tensor dmensonalty reducton have the advantages of hghly computng effcency. Meanwhle, there are many expermental results whch ndcate the two-dmenson method s obvously better than one-dmensonal method n some datasets. Recently, the representaton method of the two dmensonlzaton and tensor model has been wdely appled n the felds of the machne learnng, pattern recognton and computer recognton. Tensor Lnear Dscrmnate Analyss,The tensor lnear dscrmnate algorthm tres to expand the dstance between classes. Meanwhle, t tres to shrnk the dstance wthn a class. There are some smlar between the thought of the tensor lnear dscrmnate analyss and the tradtonal lnear dscrmnate analyss. The dfferences s that the tradtonal lnear dscrmnate analyss method s only to seek a projecton vector x as n-dmensonal sample; N s the total sample; X belongs to the selecton of classes s = ( = 1,, N s ) ; N S s the sample of the sth categores; c s the number of categores for all samples. However, the tensor lnear dscrmnate analyss method s to seek and obtan the process of a seres of projecton matrx. Then t transforms the tensor sample nto the tensor data of the low dmenson. Meanwhle t would maxmze the class dstance and mnmze the wthn class dstance. The tensor lnear dscrmnate analyss method s to deal wth samples and extract the features of samples n the tensor feld. Ths can be consdered as an extenson of the tradtonal lnear dscrmnate analyss method n the tensor subspace. Compared wth vector, the tensor subspace effectvely retaned the sample nformaton and made full use of the collectng nformaton. At the same tme, the tensor lnear dscrmnate analyss method overcomes the ssue of the small sample sze n the tradtonal lnear dscrmnate method and mproves the learnng performance Manfold Learnng and Localzed Dmensonalty Reducton The manfold learnng has become a new hot research n machne learnng feld. Ths s happened n 000 whch s the year of the magazne scence publshed two classc artcles about the sometrc mappng (ISOMAP) and locally lnear embeddng (LLE). The manfoldng learnng s another way to obtan nonlnear dmensonalty reducton except for the kernel method. One of the basc assumptons of the manfold learnng s to dstrbute samples on a potental manfold. Thus, the nternal dmensonalty of the manfold s generally not hgh even though the nput space s the hgh-dmensonal space. After the emergence of Isomer and LLE method, there have been a number of other manfold learnng algorthms. These algorthms nvolved maxmum varance extenson, Laplace feature mappng and local tangent space algnment so on. Compared wth the tradtonal lnear dmensonalty reducton algorthm, the manfold learnng can effectvely solve the problem of the nonlnear. However, most manfold algorthms stll have some ssues about the low-effcency of computng and dffcult extendng of the test samples. The followng algorthm has been mproved for above problems respectvely. The locally preservng projecton s proposed by He (X. He) and other researchers s one of the most representatves. As a matter of fact, the locally preservng projecton s the lberalzed verson of the Laplace feature mappng. There are many dmensonalty reducton algorthms of the localzaton besdes the localty preservng projectons. These methods all utlzed the thoughts of the local preservaton. Wth the ntroducton of the class nterval thoughts of the support vector machne, the dmensonalty reducton algorthm of the nearest neghbor dscrmnate analyss was proposed. The comparson of the nearest neghbor dscrmnate analyss and localty preservng projecton shows that the nearest neghbor dscrmnate analyss not only can ntegrate category nformaton but also can obtan the dmensonalty reducton of the optmal dscrmnate ablty. Recently, a relevant typcal example was proposed n the scentfc feld. Yan (S. Yan) and other researchers proposed a general framework of the map embeddng. Most of the above dmensonalty reducton methods was nvolved n ths framework. The constructon and the selecton of parameters of a map s the key step n the dmenson redacton methods of the map embeddng. 4. The Summary and Prospect of the Dmenson Reducton Methods Undoubtfully, data dmensonalty reducton s an effectve tool for the work of the data dggng. That nformaton whch s hdden by the hgh-dmensonal data provdes a possblty for solvng the scentfc problems. Ths possblty s dependng on the mprovement and development of the dmenson reducton methods. The dmenson reducton method can obtan an effectvely low-dmensonal representaton and a way of the essental cha- 1

7 racterstcs of reflectng the orgnal data. Then the method of processng low-dmensonal data can be used. Ths can avod data dsaster. In terms wth the nature of the data, the method of dmenson reducton can be dvded nto two categores. One category s Lnear Dmensonalty Reducton. The other s Nonlnear Dmensonalty Reducton. In fact, the Lnear Dmensonalty Reducton s based on the dfferent crterson of optmzaton and to seek the optmal lnear model. Ths s totally dfferent from the Nonlnear Dmensonalty Reducton. However, the data of the practcal applcaton s usually not the lnear combnaton wthn features. Consequently, t trggered the study of the manfold learnng n the feld of the nonlnear dmensonalty reducton. The manfold learnng method commonly can be dvded nto three categores. These three categores are the global comparson of the lnear model, the nonlnear method of preservng local property and the nonlnear method of preservng global property. The method of the global comparson of the lnear model majorly refers to LLC. Whle the nonlnear method of preservng local property ncludes LLE, Hessan LLE, Palladan Egen map and LTSA. The nonlnear method of preservng global property nvolves the nonlnear dmenson reducton of the multdmensonal scalng of based on kernel (Kernel PCA, Fsher Analyss), Isomap and Dffuson map so on. The man dfferences between the global method and the local method le n the local structure and the way of embeddng. The dmenson reducton method was wdely welcomed n many felds snce t was proposed. These felds are the face recognton feld and cluster analyss feld so on. Over the past years, the dmenson reducton method has been mproved. However, there are some lmtatons n realty. These lmtatons especally presented n the manfold method. The frst lmtaton s that t s dffcult to obtan the mappng relatonshp from the hgh-dmensonal space to the low-dmensonal space durng the process of dmensonalty reducton. The second lmtaton s the dmenson reducton problem of processng dmenson jump. Thrdly, t s dffcult to get an effectve study method of the manfold as the data collecton s relatvely sparse. Fnally, t s lack of the studes n mprovng the effcency of handlng large amounts of data after the combnaton of the manfold algorthm and parallel computng. References [1] Donoho, L. (000) Hgh Dmensonal Data Analyss: the Curses and Blessngs of Dmensonalty. Present Data Amercan Mathematcs Socety Conference. [] Lu, T. (005) Dmenson Reducton Theory and Applcaton of the Hgh Dmensonal Data. 5-31, [3] XaoSheng, Y. and Nng, Z. (007) The Method Research of the Dmenson Reducton. Informaton Scence 8th. [4] SongCan, C. and DaoQang, Z. (009) The dmenson Reducton of the Hgh Dmensonal Data. Chna Computer Federaton Informaton 8th. [5] Fan, J.Q. and Fan, Y.Y. (008) Hgh Dmensonal Classfcaton Usng Features Annealed Independent Rules. Annals of Statstcs, January 3. [6] Van der Maaten (009) Dmensonalty Reducton: A Comparatve Revew. Journal of Machne, 1-. [7] Demers, D. and Cottrell, G. Non-Lnear Dmensonalty Reducton. Dept. of Computer Scence & Engr, September, 67. [8] ZhongBao, L., Guangzhou, P. and WengJuan, Z. (013) The Manfold Dscrmnate Analyss. Journal of Electroncs & Informaton Technology 9th. [9] DeWeng, H., JanZhou, X., JunSong, Y. and Tan, Z. (007) The Analyss and Applcaton of the Nonlnear Manfold Learnng. Progress n Natural Scence 8th. [10] Wood, F. (009) Prncpal Component Analyss. December 68. [11] Ma, R., Song, Y.X. and Wang, J.X. (008) Mult-Manfold Learnng Usng Locally Lnear Embeddng Nonlnear Dmensonalty Reducton. Journal of Tsnghua Unversty, [1] Rowesst, S. (000) Nonlnear Dmensonalty Reducton by Locally Lnear Embeddng. Scence, December. 13

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

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

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

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

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

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

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

An Indian Journal FULL PAPER ABSTRACT KEYWORDS. Trade Science Inc.

An Indian Journal FULL PAPER ABSTRACT KEYWORDS. Trade Science Inc. [Type text] [Type text] [Type text] ISSN : 97-735 Volume Issue 9 BoTechnology An Indan Journal FULL PAPER BTAIJ, (9), [333-3] Matlab mult-dmensonal model-based - 3 Chnese football assocaton super league

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

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

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

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

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

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

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

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc.

BioTechnology. An Indian Journal FULL PAPER. Trade Science Inc. [Type text] [Type text] [Type text] ISSN : 0974-74 Volume 0 Issue BoTechnology 04 An Indan Journal FULL PAPER BTAIJ 0() 04 [684-689] Revew on Chna s sports ndustry fnancng market based on market -orented

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

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

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

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

A Semi-parametric Regression Model to Estimate Variability of NO 2

A Semi-parametric Regression Model to Estimate Variability of NO 2 Envronment and Polluton; Vol. 2, No. 1; 2013 ISSN 1927-0909 E-ISSN 1927-0917 Publshed by Canadan Center of Scence and Educaton A Sem-parametrc Regresson Model to Estmate Varablty of NO 2 Meczysław Szyszkowcz

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

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

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

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration

Improvement of Spatial Resolution Using BlockMatching Based Motion Estimation and Frame. Integration Improvement of Spatal Resoluton Usng BlockMatchng Based Moton Estmaton and Frame Integraton Danya Suga and Takayuk Hamamoto Graduate School of Engneerng, Tokyo Unversty of Scence, 6-3-1, Nuku, Katsuska-ku,

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

Evaluation of the application of BIM technology based on PCA - Q Clustering Algorithm and Choquet Integral

Evaluation of the application of BIM technology based on PCA - Q Clustering Algorithm and Choquet Integral IETI Transactons on Busness and Management Scences, 2016, Volume 1, Issue 1, 47-55. http://www.et.net/tc An Internatonal Open Access Journal Evaluaton of the applcaton of BIM technology based on PCA -

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

Laplacian Eigenmap for Image Retrieval

Laplacian Eigenmap for Image Retrieval Laplacan Egenmap for Image Retreval Xaofe He Partha Nyog Department of Computer Scence The Unversty of Chcago, 1100 E 58 th Street, Chcago, IL 60637 ABSTRACT Dmensonalty reducton has been receved much

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

A new segmentation algorithm for medical volume image based on K-means clustering

A new segmentation algorithm for medical volume image based on K-means clustering Avalable onlne www.jocpr.com Journal of Chemcal and harmaceutcal Research, 2013, 5(12):113-117 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCRC5 A new segmentaton algorthm for medcal volume mage based

More information

Positive Semi-definite Programming Localization in Wireless Sensor Networks

Positive Semi-definite Programming Localization in Wireless Sensor Networks Postve Sem-defnte Programmng Localzaton n Wreless Sensor etworks Shengdong Xe 1,, Jn Wang, Aqun Hu 1, Yunl Gu, Jang Xu, 1 School of Informaton Scence and Engneerng, Southeast Unversty, 10096, anjng Computer

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

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

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures

A Novel Adaptive Descriptor Algorithm for Ternary Pattern Textures A Novel Adaptve Descrptor Algorthm for Ternary Pattern Textures Fahuan Hu 1,2, Guopng Lu 1 *, Zengwen Dong 1 1.School of Mechancal & Electrcal Engneerng, Nanchang Unversty, Nanchang, 330031, Chna; 2. 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

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

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

Semi-Supervised Discriminant Analysis Based On Data Structure

Semi-Supervised Discriminant Analysis Based On Data Structure IOSR Journal of Computer Engneerng (IOSR-JCE) e-issn: 2278-0661,p-ISSN: 2278-8727, Volume 17, Issue 3, Ver. VII (May Jun. 2015), PP 39-46 www.osrournals.org Sem-Supervsed Dscrmnant Analyss Based On Data

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

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

Design of Structure Optimization with APDL

Design of Structure Optimization with APDL Desgn of Structure Optmzaton wth APDL Yanyun School of Cvl Engneerng and Archtecture, East Chna Jaotong Unversty Nanchang 330013 Chna Abstract In ths paper, the desgn process of structure optmzaton wth

More information

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation

Quality Improvement Algorithm for Tetrahedral Mesh Based on Optimal Delaunay Triangulation Intellgent Informaton Management, 013, 5, 191-195 Publshed Onlne November 013 (http://www.scrp.org/journal/m) http://dx.do.org/10.36/m.013.5601 Qualty Improvement Algorthm for Tetrahedral Mesh Based on

More information

Supervised Nonlinear Dimensionality Reduction for Visualization and Classification

Supervised Nonlinear Dimensionality Reduction for Visualization and Classification IEEE Transactons on Systems, Man, and Cybernetcs Part B: Cybernetcs 1 Supervsed Nonlnear Dmensonalty Reducton for Vsualzaton and Classfcaton Xn Geng, De-Chuan Zhan, and Zh-Hua Zhou, Member, IEEE Abstract

More information

The Comparison of Calibration Method of Binocular Stereo Vision System Ke Zhang a *, Zhao Gao b

The Comparison of Calibration Method of Binocular Stereo Vision System Ke Zhang a *, Zhao Gao b 3rd Internatonal Conference on Materal, Mechancal and Manufacturng Engneerng (IC3ME 2015) The Comparson of Calbraton Method of Bnocular Stereo Vson System Ke Zhang a *, Zhao Gao b College of Engneerng,

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

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

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

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes

R s s f. m y s. SPH3UW Unit 7.3 Spherical Concave Mirrors Page 1 of 12. Notes SPH3UW Unt 7.3 Sphercal Concave Mrrors Page 1 of 1 Notes Physcs Tool box Concave Mrror If the reflectng surface takes place on the nner surface of the sphercal shape so that the centre of the mrror bulges

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

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

Principal Component Inversion

Principal Component Inversion Prncpal Component Inverson Dr. A. Neumann, H. Krawczyk German Aerospace Centre DLR Remote Sensng Technology Insttute Marne Remote Sensng Prncpal Components - Propertes The Lnear Inverson Algorthm Optmsaton

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

A Workflow for Spatial Uncertainty Quantification using Distances and Kernels

A Workflow for Spatial Uncertainty Quantification using Distances and Kernels A Workflow for Spatal Uncertanty Quantfcaton usng Dstances and Kernels Célne Schedt and Jef Caers Stanford Center for Reservor Forecastng Stanford Unversty Abstract Assessng uncertanty n reservor performance

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

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

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

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

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

The Research of Ellipse Parameter Fitting Algorithm of Ultrasonic Imaging Logging in the Casing Hole

The Research of Ellipse Parameter Fitting Algorithm of Ultrasonic Imaging Logging in the Casing Hole Appled Mathematcs, 04, 5, 37-3 Publshed Onlne May 04 n ScRes. http://www.scrp.org/journal/am http://dx.do.org/0.436/am.04.584 The Research of Ellpse Parameter Fttng Algorthm of Ultrasonc Imagng Loggng

More information

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z.

TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS. Muradaliyev A.Z. TECHNIQUE OF FORMATION HOMOGENEOUS SAMPLE SAME OBJECTS Muradalyev AZ Azerbajan Scentfc-Research and Desgn-Prospectng Insttute of Energetc AZ1012, Ave HZardab-94 E-mal:aydn_murad@yahoocom Importance of

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

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

The Study of Remote Sensing Image Classification Based on Support Vector Machine

The Study of Remote Sensing Image Classification Based on Support Vector Machine Sensors & Transducers 03 by IFSA http://www.sensorsportal.com The Study of Remote Sensng Image Classfcaton Based on Support Vector Machne, ZHANG Jan-Hua Key Research Insttute of Yellow Rver Cvlzaton and

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

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

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

A Study on Clustering for Clustering Based Image De-Noising

A Study on Clustering for Clustering Based Image De-Noising Journal of Informaton Systems and Telecommuncaton, Vol. 2, No. 4, October-December 2014 196 A Study on Clusterng for Clusterng Based Image De-Nosng Hossen Bakhsh Golestan* Department of Electrcal Engneerng,

More information

SELECTION OF THE NUMBER OF NEIGHBOURS OF EACH DATA POINT FOR THE LOCALLY LINEAR EMBEDDING ALGORITHM

SELECTION OF THE NUMBER OF NEIGHBOURS OF EACH DATA POINT FOR THE LOCALLY LINEAR EMBEDDING ALGORITHM ISSN 392 24X INFORMATION TECHNOLOGY AND CONTROL, 2007, Vol.36, No.4 SELECTION OF THE NUMBER OF NEIGHBOURS OF EACH DATA POINT FOR THE LOCALLY LINEAR EMBEDDING ALGORITHM Rasa Karbauskatė,2, Olga Kurasova,2,

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

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

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

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 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

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

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

Wishing you all a Total Quality New Year!

Wishing you all a Total Quality New Year! Total Qualty Management and Sx Sgma Post Graduate Program 214-15 Sesson 4 Vnay Kumar Kalakband Assstant Professor Operatons & Systems Area 1 Wshng you all a Total Qualty New Year! Hope you acheve Sx sgma

More information

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints

Sum of Linear and Fractional Multiobjective Programming Problem under Fuzzy Rules Constraints Australan Journal of Basc and Appled Scences, 2(4): 1204-1208, 2008 ISSN 1991-8178 Sum of Lnear and Fractonal Multobjectve Programmng Problem under Fuzzy Rules Constrants 1 2 Sanjay Jan and Kalash Lachhwan

More information

Available online at ScienceDirect. Procedia Environmental Sciences 26 (2015 )

Available online at   ScienceDirect. Procedia Environmental Sciences 26 (2015 ) Avalable onlne at www.scencedrect.com ScenceDrect Proceda Envronmental Scences 26 (2015 ) 109 114 Spatal Statstcs 2015: Emergng Patterns Calbratng a Geographcally Weghted Regresson Model wth Parameter-Specfc

More information

Two-Dimensional Supervised Discriminant Projection Method For Feature Extraction

Two-Dimensional Supervised Discriminant Projection Method For Feature Extraction Appl. Math. Inf. c. 6 No. pp. 8-85 (0) Appled Mathematcs & Informaton cences An Internatonal Journal @ 0 NP Natural cences Publshng Cor. wo-dmensonal upervsed Dscrmnant Proecton Method For Feature Extracton

More information

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems

A Unified Framework for Semantics and Feature Based Relevance Feedback in Image Retrieval Systems A Unfed Framework for Semantcs and Feature Based Relevance Feedback n Image Retreval Systems Ye Lu *, Chunhu Hu 2, Xngquan Zhu 3*, HongJang Zhang 2, Qang Yang * School of Computng Scence Smon Fraser Unversty

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

Intra-Parametric Analysis of a Fuzzy MOLP

Intra-Parametric Analysis of a Fuzzy MOLP Intra-Parametrc Analyss of a Fuzzy MOLP a MIAO-LING WANG a Department of Industral Engneerng and Management a Mnghsn Insttute of Technology and Hsnchu Tawan, ROC b HSIAO-FAN WANG b Insttute of Industral

More information

Load Balancing for Hex-Cell Interconnection Network

Load Balancing for Hex-Cell Interconnection Network Int. J. Communcatons, Network and System Scences,,, - Publshed Onlne Aprl n ScRes. http://www.scrp.org/journal/jcns http://dx.do.org/./jcns.. Load Balancng for Hex-Cell Interconnecton Network Saher Manaseer,

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

High resolution 3D Tau-p transform by matching pursuit Weiping Cao* and Warren S. Ross, Shearwater GeoServices

High resolution 3D Tau-p transform by matching pursuit Weiping Cao* and Warren S. Ross, Shearwater GeoServices Hgh resoluton 3D Tau-p transform by matchng pursut Wepng Cao* and Warren S. Ross, Shearwater GeoServces Summary The 3D Tau-p transform s of vtal sgnfcance for processng sesmc data acqured wth modern wde

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

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

Review of approximation techniques

Review of approximation techniques CHAPTER 2 Revew of appromaton technques 2. Introducton Optmzaton problems n engneerng desgn are characterzed by the followng assocated features: the objectve functon and constrants are mplct functons evaluated

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

Journal of Chemical and Pharmaceutical Research, 2014, 6(6): Research Article. A selective ensemble classification method on microarray data

Journal of Chemical and Pharmaceutical Research, 2014, 6(6): Research Article. A selective ensemble classification method on microarray data Avalable onlne www.ocpr.com Journal of Chemcal and Pharmaceutcal Research, 2014, 6(6):2860-2866 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCPRC5 A selectve ensemble classfcaton method on mcroarray

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

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

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation

An Iterative Solution Approach to Process Plant Layout using Mixed Integer Optimisation 17 th European Symposum on Computer Aded Process Engneerng ESCAPE17 V. Plesu and P.S. Agach (Edtors) 2007 Elsever B.V. All rghts reserved. 1 An Iteratve Soluton Approach to Process Plant Layout usng Mxed

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

IMAGE FUSION BASED ON EXTENSIONS OF INDEPENDENT COMPONENT ANALYSIS

IMAGE FUSION BASED ON EXTENSIONS OF INDEPENDENT COMPONENT ANALYSIS IMAGE FUSION BASED ON EXTENSIONS OF INDEPENDENT COMPONENT ANALYSIS M Chen a, *, Yngchun Fu b, Deren L c, Qanqng Qn c a College of Educaton Technology, Captal Normal Unversty, Bejng 00037,Chna - (merc@hotmal.com)

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

ISSN: International Journal of Engineering and Innovative Technology (IJEIT) Volume 1, Issue 4, April 2012

ISSN: International Journal of Engineering and Innovative Technology (IJEIT) Volume 1, Issue 4, April 2012 Performance Evoluton of Dfferent Codng Methods wth β - densty Decodng Usng Error Correctng Output Code Based on Multclass Classfcaton Devangn Dave, M. Samvatsar, P. K. Bhanoda Abstract A common way to

More information

A Clustering Algorithm for Key Frame Extraction Based on Density Peak

A Clustering Algorithm for Key Frame Extraction Based on Density Peak Journal of Computer and Communcatons, 2018, 6, 118-128 http://www.scrp.org/ournal/cc ISSN Onlne: 2327-5227 ISSN Prnt: 2327-5219 A Clusterng Algorthm for Key Frame Extracton Based on Densty Peak Hong Zhao

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