A Novel Term_Class Relevance Measure for Text Categorization
|
|
- Junior Atkinson
- 5 years ago
- Views:
Transcription
1 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 called Term_Class relevance to compute the relevancy of a term n classfyng a document nto a partcular class. The proposed measure estmates the degree of relevance of a gven term, n placng an unlabeled document to be a member of a known class, as a product of Class_Term weght and Class_Term densty; where the Class_Term weght s the rato of the number of documents of the class contanng the term to the total number of documents contanng the term and the Class_Term densty s the relatve densty of occurrence of the term n the class to the total occurrence of the term n the entre populaton. Unlke the other exstng term weghtng schemes such as TF-IDF and ts varants, the proposed relevance measure takes nto account the degree of relatve partcpaton of the term across all documents of the class to the entre populaton. To demonstrate the sgnfcance of the proposed measure expermentaton has been conducted on the 20 Newsgroups dataset. Further, the superorty of the novel measure s brought out through a comparatve analyss. Keywords: text categorzaton, term weght, term-document relevance, term_class relevance. 1. Introducton For the few decades automatc content based classfcaton of documents from huge collectons has become an actve area of research due to the fact that electronc data over the nternet has become unmanageably bg and day by day t s ncreasng exponentally. Manual, tag based classfcaton have lost ther sgnfcance because of the huge sze of the data that need to be processed and nablty of the tags n descrbng the content of the documents. Varetes of applcatons of text classfcaton whch are of current demand such as spam flterng n E-mals, classfcaton of E-Books, classfcaton of news documents, classfcaton of text data from socal networks and so on have also made the researchers to explore varous ways of analyzng and representng these data so that quck and effcent retreval and management of ths huge data can be done. 1.1 A revew of the avalable term weghtng schemes As our work focuses on proposal of a new term weghtng scheme, but not on classfcaton framework, here we consder the lterature on only dfferent term weghtng schemes. Terms are the basc nformaton unts of any text document. So, all weghtng schemes developed n the lterature measure the weght of a term n representng the content of a document [1-5]. Based on whether the membershp of the document n predefned categores s provded to
2 measure the weght of a term or not, term weghtng schemes are broadly classfed nto two classes namely, unsupervsed term weghtng schemes and supervsed term weghtng schemes. In the followng subsectons we provde a revew of both the weghtng scheme along wth the technques whch have adopted them Unsupervsed term weghtng schemes Most of the unsupervsed term weghtng schemes are from the nformaton retreval feld. These methods are very useful when the tranng documents are not labeled by ther class labels. The tradtonal term weghtng methods borrowed from IR, such as bnary, term frequency (TF), TF- IDF, and ts varous varants are unsupervsed schemes [2]. The TF-IDF proposed by Jones [6, 7] and ts varants are the most wdely used term weghtng schemes for text classfcaton. Some of the varants of TF are Raw term frequency, log(tf), log(tf+1), or log(tf)+1[1-2]. If n s the number of documents contanng the term and N s the number of documents n the collecton then, the varants of IDF are 1/n, log(1/n), log(n/n), log(n/n)+1 and log(n/n-1)[1]. In [18], a novel nverse corpus frequency (ICF) based technque s proposed whch computes the document representaton n lnear tme Supervsed term weghtng schemes Supervsed term weghtng schemes were developed especally for text categorzaton because of the fact that a supervsed knowledge on the class labels of the tranng samples s provded [1-4]. All the supervsed term weghtng schemes make use of ths class nformaton n dfferent ways. Supervsed term weghtng schemes are further classfed nto subcategores, based on whether the weght estmates relevancy of a term n preservng document content or the relevancy of a term n placng a document as a member of a class. So, t wll be more effectve to call the weghtng schemes whch are used to measure the relevance of a term n preservng the document content as term-document relevance measures and those whch can be used to measure the term relevance n categorzng a document as term_class relevance measures. Term-Document Relevance measure These measures are useful to select a dscrmnatng subset of terms for representng a document by weghtng the terms accordng to ther relevance n preservng the content of the document. These are created by replacng the IDF component of the TF-IDF scheme. Most frequently used technques to replace IDF nclude ch-square measure (X 2 ), Informaton Gan(IG), Gan Rato, Mutual Informaton(MI), Odds Rato(OR) [1-4,8-12]. From past few years, many researchers have proposed alternatve term-document relevance schemes [1, 13-16]. All these are bascally feature selecton technques used n term weghtng schemes. In [14], a comparson of corpusbased and class-based keyword selecton s proposed by usng TF-IDF as weghtng scheme. In [4], a class-ndexng-based term weghtng for automatc text classfcaton s proposed. An
3 nverse class space densty frequency ( ICS F a postve dscrmnaton on nfrequent and frequent terms. ) s used along wth TF-IDF method that provdes Term_class Relevance measures These measures compute the ablty of a term n classfyng a document as a member of a class. To the best of our knowledge, only one work of ths category has been proposed by Isa et al., [20] usng Bayes posteror probablty. Though, some works make use of Bayes probablty for representaton, they have not clearly stated the advantage of the measure n classfcaton [11, 18]. After [20], ths measure was extensvely used for term weghtng [21, 22]. The beauty of ths measure les n the fact that, nstead of computng the weght of a term n preservng the content of a document, the relevancy of the term n categorzng the document as a member of a class can be measured drectly. Whch s computed as the Bayes posteror probablty P(C/ t) for a class C and term t as gven by, where, P( t / C ) P( C ) P( C / t) Pt ( ) Total _ of _ Words _ n _ C PC ( ), Total _ of _ Words _ n _ Tranng _ Dataset Pt ( ) occurrence _ of _ t _ n _ all _ categores, and occurrence _ of _ all _ terms _ n _ all _ categores occurence _ of _ t _ n _ C P( t / C ) occurrence _ of _ all _ terms _ n _ C To make use of the complete advantage of the proposed relevance measure, Isa et al., [20] also propose a text representaton scheme whch works wth the reduced dmenson for each document at the tme of representaton tself. Ths work happened to be the very frst of ts knd n the lterature of text classfcaton where, a document s represented only wth number of dmensons equal to the number of classes n the corpus wthout any dmensonalty reducton technque appled. In ths representaton scheme, frst, a matrx F of sze m X k s created for every document where, m s the number of terms assumed to be avalable n the document and k s the number of classes. Then, every entry F(, ) of the matrx s flled by the relevancy of the correspondng term t n classfyng the correspondng document as a member of class C. Then, a feature vector
4 f of dmenson k s created as a representatve for the document where, f() s the average of relevancy of every term to a class C. It shall be carefully observed here that, a document wth any number of terms s represented wth a feature vector of dmenson equal to the number of classes n the populaton whch s very small n contrast to the feature vector that s created n any other vector space representaton scheme where the dmenson s equal to the total number of terms due to all documents of the populaton. Therefore, a great amount of dmensonalty reducton s acheved at the tme of representaton tself wthout the applcaton of any dmensonalty reducton technque. However, the classfcaton accuracy accomplshed s not of that hgh. Motvated by ths work, n ths paper we propose a novel term_class relevance measure wth the followng obectves, Explotng the complete advantage of text representaton scheme proposed by Isa et al.,[20]. Comparson of the effectveness of the proposed term_class relevance measure wth that of Bayes posteror probablty based measure. Isa et al., [20] make use of SVM as the classfer. So we are also nvestgatng the effect SVM on our proposed relevance measure and also compare t wth other avalable classfers. The rest of the paper s organzed as follows. The proposed Class_Term relevance measure s presented n the Secton 2. In secton 3, presents the results and dscusson on the expermentaton. A comparatve analyss of the proposed relevance measure wth other contemporary works s gven n the Secton 4. Fnally, secton 6 presents the concluson and future enhancements. 2. A New Term_Class Relevance Measure In ths secton, we propose a novel measure called term_class relevance measure. Term_class relevancy s defned as the ablty of a term t n classfyng a document D as a member of a class C. We begn wth ntroducng two new concepts whch decde the role of a term n a class, namely, Class_Term Weght and Class_Term Densty. Class_Term Weght: It s the relatve weght of the term wth respect to a class of nterest whch s computed by countng only those documents of the class of nterest that are contanng the term of nterest aganst that of the entre corpus. That s, the class_term weght of a term t n the class C s computed as the rato of ClassFrequency ( t, C ) to the CorpusFrequency ( t ). It s gven by the equaton below. ClassFrequency ( t, C ) Class _ TermWeght ( t, C ) CorpusFrequency ( t )
5 where, ClassFrequency ( t, C ) s the number of documents of C contanng t at least once and CorpusFrequency ( t ) once. s the number of documents of the entre corpus contanng t at least If the class_term weght of a term t wth respect to the class C s very hgh then the probablty that the document D whch contans t s most lkely a member of the class C s also hgh. Therefore, the relevancy of a term whch we call t as Term _ ClassRe levancy ( t, C ) n decdng the class of a document s drectly proportonal to the class_term weght of the term..e., Term _ Class Re levancy ( t, C ) Class _ TermWeght ( t, C ) (1) Class_Term Densty: It s the relatve densty of a term of nterest wth respect to the class of nterest. It s computed as the rato of the number of occurrences of the term n the class of nterest to that of the entre corpus. That s, the class_term densty of a term t wth respect to the class C s computed as the rato of frequency of t n C to ts frequency n the corpus. It s gven by the equaton below. Class _ TermDensty( t, C ) k TermFrequency ( t, C ) 1 TermFrequency ( t, C ) where, TermFrequency ( t, C ) s the frequency of t n the class C whch s computed as the sum of the frequences of t n every document of C as shown by the equaton below. TermFrequency ( t, C ) Frequency ( t, D ) d doc doc1 where, Frequency( t, D) s the frequency of occurrence of the term t n document D and d s the number of documents n the class C. It shall be notced that, f the class_term densty of a term t n a class C s very hgh then the probablty that a document D whch contans t s most lkely a member of the class C s also hgh. Therefore, the relevancy of a term n decdng the class of a document s drectly proportonal to the class_term densty of the term..e., Term _ Class Re levancy ( t, C ) Class _ TermDensty ( t, C ) (2) By combnng (1) and (2), the term_class relevancy s drectly proportonal to the product of the class_term weght and class_term densty of the term, Term _ Class Re levancy ( t, C ) Class _ TermWeght ( t, C )* Class _ TermDensty ( t, C ) Term _ Class Re levancy ( t, C ) c* Class _ TermWeght ( t, C )* Class _ TermDensty ( t, C )
6 .e., where, c s the proportonalty constant, whch we decde based on the class weght wth respect to the entre populaton. Class Weght ( c ): It s the weght of the th class C n the corpus whch s computed as the rato of the number of documents n C denoted by Sze _ of ( C ) to the total number of documents n the entre corpus as gven by, ClassWeght( C ) Sze _ of ( C ) k 1 Sze _ of ( C ) Where, k s the number of classes. If each class has equal number of documents, then the class-weght serves as a scalng factor n computng the relevance of a term and t ncreases or decreases the relevancy of a term to a class when the sze of the class compared to the sze of other classes s larger or smaller respectvely. Therefore, the proposed relevancy measure of a term t n placng a document D as a member of a class C s gven by the product of the three aspects namely, Class weght, Class_Term weght and Class_Term Densty as gven by the formula below. Term _ Class Re levancy ( t, C ) c* Class _ TermWeght ( t, C )* Class _ TermDensty ( t, C ) The man advantages of the proposed term_class relevancy measure are as follows, It drectly computes the relevancy of the term wth respect to a class of nterest; whch can tself be used as a clue to dentfy the possble class to whch a document may belong wthout the need of a classfer. The measure uses class as well as corpus nformaton together as opposed to the conventonal TF-IDF scheme, whch utlzes the document frequency from only the corpus. It shall be observed that, the relevancy of a term to a class s hgh only f the three factors class_term weght, class_term densty and class_weght are hgh. Ths helps n properly decdng the weght of a term wthout any bas towards a partcular class, whch n turn helps n decdng the class for a classfer. Once the term_class relevance of all terms of the tranng set of documents s computed wth respect to every class present, each tranng document s then represented usng the representaton scheme proposed by Isa et al., [20] as explaned n secton A document s frst represented as a matrx of sze, where, m s the number of terms assumed to be avalable n the document and k s the number of classes. Then, every entry of the matrx s flled by the relevancy of the correspondng term t wth respect to the class C. Then, a feature vector f of
7 dmenson k s created as a representatve for the document where, f() s the average relevancy of all terms wth respect to a class C. The feature matrx of sze thus created for the n tranng documents s used for learnng process. A smlar vector of k dmenson s created for the test documents and gven to the learnng algorthm or a classfer for labelng. The process of tranng and testng the classfers s explaned n the next secton. 3. Classfcaton wth SVM and k-nn classfers To evaluate the applcablty of the proposed term_class relevance measure, we make use SVM as learnng algorthm to perform classfcaton because of ts good generalzaton ablty. Moreover, the tranng burden for SVM s very less even though, the tme requred for tranng s drectly proportonal to the tranng dataset, because the representatve feature vectors are of dmenson equal to the number of classes only. So, to test the effectveness of the proposed relevance measure we have expermented wth the SVM classfer wth Lnear, Gaussan radal bass functon (RBF) and Polynomal kernels. We consder the 20 Newsgroups data set for our expermentaton. It conssts of approxmately 20,000 newsgroup documents consstng 20 classes wth each class bearng nearly equal number of samples. It has become a popular data set for text classfcaton and clusterng applcatons. Some of the documents are closely related to each other whle others are hghly unrelated. We conduct experments wth varous proporton of tranng set to valdate the performance of the proposed relevancy measure. Fg 1 shows the overall classfcaton accuracy of the system wth varous percentages of tranng samples usng SVM classfer wth dfferent kernels. Fg 2 shows the precson of the SVM classfer wth dfferent kernels and Recall s shown n Fg 3. In Fg 4, the overall F- measure s presented. It can be observed from the fgures (1-4) that, the SVM classfer wth RBF kernel s workng well when compared to the other kernels. The results are also presented graphcally n fgures below. 90 A c c u r a c y percentage of tranng Lnear RBF Polynomal
8 Fgure 1. Overall accuracy of classfcaton wth Lnear, RBF and Polynomal kernels 90 P r e c s o n Percentage of Tranng Lnear RBF Polynomal Fgure 2. precson of the SVM classfer wth Lnear, RBF and Polynomal kernels R e c a l l Percentage of Tranng Lnear RBF Polynomal Fgure 3. Recall of the SVM classfer wth Lnear, RBF and Polynomal kernels F - M e a s u r e Lnear RBF Polynomal
9 F-measure Fgure 4. F-measure of the SVM classfer wth Lnear, RBF and Polynomal kernels Further, the k-nn classfer s also adapted to test the proposed method because of ts smplcty n classfcaton. We performed the expermentaton wth varous values of k from 1 to 20 and the performance of the classfer was hgh for k=10. Table 2 shows the results of k-nn classfer for k=10 and a comparson wth the best results of SVM s also gven. Table 2. Results of SVM wth RBF kernel and k-nn wth k=10 % of Accuracy Precson Recall F-measure Trang k-nn SVM k-nn SVM k-nn SVM k-nn SVM To compare the class-wse performance of each classfer we show the varaton of F-measure vs. class n Fgure 5 and 6. Fgure 5, shows the values of F-measure vs. each class usng k-nn classfer wth k=10 and 10 percent of tranng. It can be notced that, the performance s relatvely low for classes 2, 3, 4, 7, 13 and 20. Further, the F-measure of SVM classfer vs. each class wth RBF kernel and 10 % of tranng s shown n Fgure 6. Though, the results of SVM are poor when compared to k-nn, SVM also has shown relatvely low performance for the same classes as n the case of k-nn Class Number
10 F-measure Fgure 5. Classfcaton performance vs. class for k-nn classfer wth 10 % tranng Class Number Fgure 6. Classfcaton performance vs. class for SVM classfer wth RBF kernel and 10 % tranng 4. Comparatve Analyss In ths secton, we provde a quanttatve comparatve analyss of the proposed term_class relevance measure wth the results of Isa et al.,[20] n Table 3. The results correspondng to [20] have been extracted drectly from the paper as the representaton scheme s same n both the works and they also have provded the results on the same 20Newsgroups dataset usng only SVM classfer wth dfferent kernels. We can notce from the Table 3 that, the proposed term_class relevance measure outperforms the measure used by Isa et al.,[20]. Along wth SVM, we compare the results usng results of k-nn classfer wth k=10. It can also be notced from the Table 3 that, k-nn classfer wth k=10 s showng enhanced results when compared to SVM wth all the kernels for both the relevance measures. So, we recommend usng k-nn as the classfer for better classfcaton performance. Table 3. Comparson of Results of the proposed method wth the work of Isa et al.,[20]. Percentage of Tranng Results from [20] wth SVM Results of Proposed Method SVM Lnear RBF Polynomal Lnear RBF Polynomal k-nn
11 Concluson In ths paper, a novel term_class relevance measure to compute the relevance of a term n classfyng an unknown document as a member of a partcular class s proposed. The proposed term_class relevance measure s a product of three aspects namely class_term weght, class_term densty and class_weght. Experments are conducted on 20 Newsgroups dataset usng the SVM and k-nn classfers. An effectve text representaton scheme whch allows representaton of text documents n reduced dmenson s adapted to test the proposed term_class relevance measure. The comparatve analyss of the results of the proposed work wth the other contemporary research works shows the superorty of the proposed term_class relevance measure. References 1. Lan, M., Tan, C. L., Su. J., and Lu, Y Supervsed and Tradtonal Term Weghtng Methods for Automatc Text Categorzaton. IEEE Transactons on Pattern Analyss and Machne Intellgence, Volume: 31 (4), pp G. Salton and C. Buckley Term-Weghtng Approaches n Automatc Text Retreval, Informaton Processng and Management, vol. 24(5), pp Debole F, Sebastan. F Supervsed Term Weghtng for Automated Text Categorzaton. Proceedngs of the 2003 ACM symposum on appled computng, pp Ren F, Sohrab M. G., Class-ndexng-based term weghtng for automatc text classfcaton. Informaton Scences 236 (2013) Harsh B. S., Guru D. S., and Manunath. S. (2010). Representaton and Classfcaton of Text Documents: A Bref Revew. IJCA Specal Issue on Recent Trends n Image Processng and Pattern Recognton RTIPPR, pp K. S. Jones, A statstcal nterpretaton of term specfcty and ts applcaton n retreval, Journal of Documentaton, Vol. 28, pp K. S. Jones, A statstcal nterpretaton of term specfcty and ts applcaton n retreval, Journal of Documentaton, Vol. 60, pp Altınçay H, Erenel Z., Analytcal evaluaton of term weghtng schemes for text categorzaton. Pattern Recognton Letters Vol. 31, pp Lu, Y., Loh, H.T., Sun, A., Imbalanced text classfcaton: A term weghtng approach. Expert Systems wth Applcatons 36, Mladenc, D., Grobelnk, M., Feature selecton on herarchy of web documents. Decson Support Syst. 35 (1), Sebastan, F., Machne learnng n automated text categorzaton. ACM Comput. Surveys 34 (1), 1 47
12 12. Yang, Y., Pedersen, J.O., A comparatve study on feature selecton n text categorzaton. In: Proc. ICML 97, 14th Internat. Conf. on Machne Learnng. Morgan Kaufmann Publshers, San Francsco, US, pp Lu, H., Yu, L., Toward ntegratng feature selecton algorthms for classfcaton and clusterng. IEEE Trans. Knowledge Data Eng. 17 (4), Ozgur, A., Ozgur, L., Gungor, T., Text categorzaton wth class-based and corpusbased keyword selecton. In: Proc. 20th Internat. Symp. on Computer and Informaton Scences. Lecture Notes n Computer Scence, vol. 3733, Sprnger-Verlag, pp Tsa, R.T., Hung, H., Da, H., Ln, Y., Hsu, W., Explotng lkely-postve and unlabeled data to mprove the dentfcaton of proten proten nteracton artcles. BMC Bonform Wang, D, Zhang, H., Inverse-Category-Frequency Based Supervsed Term Weghtng Schemes for Text Categorzaton. Journal of Informaton Scence and Engneerng Vol 29, pp Reed, J, W., Jao, Y., Potok T, E., Klump, B, A., Elmore, M, T., and Hurson, A, R., TF-ICF: A New Term Weghtng Scheme for Clusterng Dynamc Data Streams. 5th Internatonal Conference on Machne Learnng and Applcatons. pp IEEE Computer Socety Washngton 18. Fuhr, N., Hartmann, S., Lustg, G., Schwantner, M., Tzeras, K., Darmstadt, T. H., et al. (1991). AIR/X A rule-based multstage ndexng system for large subect felds. In: Proceedngs of the proceedngs of RIAO(pp ) 19. P. Soucy and G.W. Mneau, Beyond tfdf Weghtng for Text Categorzaton n the Vector Space Model, Proc. Int l Jont Conf. Artfcal Intellgence, pp , Isa, D., Lee, L. H., Kallman, V. P., and Ra Kumar, R Text document preprocessng wth the Bayes formula for classfcaton usng the support vector machne. IEEE Transactons on Knowledge and Data Engneerng. Vol. 20, pp Isa, D., Kallman, V. P., Lee, L. H., Usng the self-organzng map for clusterng of text documents. Expert Systems wth Applcatons. Vol. 36, pp Guru D. S., Harsh B. S., and Manunath. S Symbolc representaton of text documents. In Proceedngs of Thrd Annual ACM Bangalore Conference. do /
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 informationThe 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 informationContent 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 informationFeature 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 informationMachine 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 informationLearning 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 informationUB at GeoCLEF Department of Geography Abstract
UB at GeoCLEF 2006 Mguel E. Ruz (1), Stuart Shapro (2), June Abbas (1), Slva B. Southwck (1) and Davd Mark (3) State Unversty of New York at Buffalo (1) Department of Lbrary and Informaton Studes (2) Department
More informationCluster 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 informationExperiments in Text Categorization Using Term Selection by Distance to Transition Point
Experments n Text Categorzaton Usng Term Selecton by Dstance to Transton Pont Edgar Moyotl-Hernández, Héctor Jménez-Salazar Facultad de Cencas de la Computacón, B. Unversdad Autónoma de Puebla, 14 Sur
More informationImpact of a New Attribute Extraction Algorithm on Web Page Classification
Impact of a New Attrbute Extracton Algorthm on Web Page Classfcaton Gösel Brc, Banu Dr, Yldz Techncal Unversty, Computer Engneerng Department Abstract Ths paper ntroduces a new algorthm for dmensonalty
More informationParallelism 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 informationPruning Training Corpus to Speedup Text Classification 1
Prunng Tranng Corpus to Speedup Text Classfcaton Jhong Guan and Shugeng Zhou School of Computer Scence, Wuhan Unversty, Wuhan, 430079, Chna hguan@wtusm.edu.cn State Key Lab of Software Engneerng, Wuhan
More informationTsinghua 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 informationBOOSTING 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 informationPerformance Evaluation of Information Retrieval Systems
Why System Evaluaton? Performance Evaluaton of Informaton Retreval Systems Many sldes n ths secton are adapted from Prof. Joydeep Ghosh (UT ECE) who n turn adapted them from Prof. Dk Lee (Unv. of Scence
More informationX- 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 informationIssues and Empirical Results for Improving Text Classification
Issues and Emprcal Results for Improvng Text Classfcaton Youngoong Ko 1 and Jungyun Seo 2 1 Dept. of Computer Engneerng, Dong-A Unversty, 840 Hadan 2-dong, Saha-gu, Busan, 604-714, Korea yko@dau.ac.kr
More informationFeature Selection for Natural Language Call Routing Based on Self-Adaptive Genetic Algorithm
IOP Conference Seres: Materals Scence and Engneerng PAPER OPEN ACCESS Feature Selecton for Natural Language Call Routng Based on Self-Adaptve Genetc Algorthm To cte ths artcle: A Koromyslova et al 017
More informationDetection 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 informationBAYESIAN 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 informationUsing an Automatic Weighted Keywords Dictionary for Intelligent Web Content Filtering
Journal of Advances n Computer Research Quarterly pissn: 2345-606x eissn: 2345-6078 Sar Branch, Islamc Azad Unversty, Sar, I.R.Iran (Vol. 6, No. 1, February 2015), Pages: 101-114 www.jacr.ausar.ac.r Usng
More informationFEATURE 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 informationMULTISPECTRAL 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 informationMachine Learning: Algorithms and Applications
14/05/1 Machne Learnng: Algorthms and Applcatons Florano Zn Free Unversty of Bozen-Bolzano Faculty of Computer Scence Academc Year 011-01 Lecture 10: 14 May 01 Unsupervsed Learnng cont Sldes courtesy of
More informationS1 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 informationInvestigating the Performance of Naïve- Bayes Classifiers and K- Nearest Neighbor Classifiers
Journal of Convergence Informaton Technology Volume 5, Number 2, Aprl 2010 Investgatng the Performance of Naïve- Bayes Classfers and K- Nearest Neghbor Classfers Mohammed J. Islam *, Q. M. Jonathan Wu,
More informationClassifying 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 informationA 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 informationProblem 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 informationNUMERICAL 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 informationCAN COMPUTERS LEARN FASTER? Seyda Ertekin Computer Science & Engineering The Pennsylvania State University
CAN COMPUTERS LEARN FASTER? Seyda Ertekn Computer Scence & Engneerng The Pennsylvana State Unversty sertekn@cse.psu.edu ABSTRACT Ever snce computers were nvented, manknd wondered whether they mght be made
More informationClassifier 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 informationNovel 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 informationQuery Clustering Using a Hybrid Query Similarity Measure
Query clusterng usng a hybrd query smlarty measure Fu. L., Goh, D.H., & Foo, S. (2004). WSEAS Transacton on Computers, 3(3), 700-705. Query Clusterng Usng a Hybrd Query Smlarty Measure Ln Fu, Don Hoe-Lan
More informationAn 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 informationAvailable online at Available online at Advanced in Control Engineering and Information Science
Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced
More informationGender Classification using Interlaced Derivative Patterns
Gender Classfcaton usng Interlaced Dervatve Patterns Author Shobernejad, Ameneh, Gao, Yongsheng Publshed 2 Conference Ttle Proceedngs of the 2th Internatonal Conference on Pattern Recognton (ICPR 2) DOI
More informationKeywords - Wep page classification; bag of words model; topic model; hierarchical classification; Support Vector Machines
(IJCSIS) Internatonal Journal of Computer Scence and Informaton Securty, Herarchcal Web Page Classfcaton Based on a Topc Model and Neghborng Pages Integraton Wongkot Srura Phayung Meesad Choochart Haruechayasak
More informationA New Feature of Uniformity of Image Texture Directions Coinciding with the Human Eyes Perception 1
A New Feature of Unformty of Image Texture Drectons Concdng wth the Human Eyes Percepton Xng-Jan He, De-Shuang Huang, Yue Zhang, Tat-Mng Lo 2, and Mchael R. Lyu 3 Intellgent Computng Lab, Insttute of Intellgent
More informationDeep Classification in Large-scale Text Hierarchies
Deep Classfcaton n Large-scale Text Herarches Gu-Rong Xue Dkan Xng Qang Yang 2 Yong Yu Dept. of Computer Scence and Engneerng Shangha Jao-Tong Unversty {grxue, dkxng, yyu}@apex.sjtu.edu.cn 2 Hong Kong
More informationClassification 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 informationA 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 informationA PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION
1 THE PUBLISHING HOUSE PROCEEDINGS OF THE ROMANIAN ACADEMY, Seres A, OF THE ROMANIAN ACADEMY Volume 4, Number 2/2003, pp.000-000 A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION Tudor BARBU Insttute
More informationUsing 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 informationA Fast Content-Based Multimedia Retrieval Technique Using Compressed Data
A Fast Content-Based Multmeda Retreval Technque Usng Compressed Data Borko Furht and Pornvt Saksobhavvat NSF Multmeda Laboratory Florda Atlantc Unversty, Boca Raton, Florda 3343 ABSTRACT In ths paper,
More informationUsing Ambiguity Measure Feature Selection Algorithm for Support Vector Machine Classifier
Usng Ambguty Measure Feature Selecton Algorthm for Support Vector Machne Classfer Saet S.R. Mengle Informaton Retreval Lab Computer Scence Department Illnos Insttute of Technology Chcago, Illnos, U.S.A
More informationAn Improvement to Naive Bayes for Text Classification
Avalable onlne at www.scencedrect.com Proceda Engneerng 15 (2011) 2160 2164 Advancen Control Engneerngand Informaton Scence An Improvement to Nave Bayes for Text Classfcaton We Zhang a, Feng Gao a, a*
More informationDetermining 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 informationReliable Negative Extracting Based on knn for Learning from Positive and Unlabeled Examples
94 JOURNAL OF COMPUTERS, VOL. 4, NO. 1, JANUARY 2009 Relable Negatve Extractng Based on knn for Learnng from Postve and Unlabeled Examples Bangzuo Zhang College of Computer Scence and Technology, Jln Unversty,
More informationClustering of Words Based on Relative Contribution for Text Categorization
Clusterng of Words Based on Relatve Contrbuton for Text Categorzaton Je-Mng Yang, Zh-Yng Lu, Zhao-Yang Qu Abstract Term clusterng tres to group words based on the smlarty crteron between words, so that
More informationSHAPE 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 informationOutline. 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 informationRecommended Items Rating Prediction based on RBF Neural Network Optimized by PSO Algorithm
Recommended Items Ratng Predcton based on RBF Neural Network Optmzed by PSO Algorthm Chengfang Tan, Cayn Wang, Yuln L and Xx Q Abstract In order to mtgate the data sparsty and cold-start problems of recommendaton
More informationResearch Article A High-Order CFS Algorithm for Clustering Big Data
Moble Informaton Systems Volume 26, Artcle ID 435627, 8 pages http://dx.do.org/.55/26/435627 Research Artcle A Hgh-Order Algorthm for Clusterng Bg Data Fanyu Bu,,2 Zhku Chen, Peng L, Tong Tang, 3 andyngzhang
More informationLecture 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 informationA Misclassification Reduction Approach for Automatic Call Routing
A Msclassfcaton Reducton Approach for Automatc Call Routng Fernando Uceda-Ponga 1, Lus Vllaseñor-Pneda 1, Manuel Montes-y-Gómez 1, Alejandro Barbosa 2 1 Laboratoro de Tecnologías del Lenguaje, INAOE, Méxco.
More informationA MODIFIED K-NEAREST NEIGHBOR CLASSIFIER TO DEAL WITH UNBALANCED CLASSES
A MODIFIED K-NEAREST NEIGHBOR CLASSIFIER TO DEAL WITH UNBALANCED CLASSES Aram AlSuer, Ahmed Al-An and Amr Atya 2 Faculty of Engneerng and Informaton Technology, Unversty of Technology, Sydney, Australa
More informationA 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 informationA Knowledge Management System for Organizing MEDLINE Database
A Knowledge Management System for Organzng MEDLINE Database Hyunk Km, Su-Shng Chen Computer and Informaton Scence Engneerng Department, Unversty of Florda, Ganesvlle, Florda 32611, USA Wth the exploson
More informationComputer Aided Drafting, Design and Manufacturing Volume 25, Number 2, June 2015, Page 14
Computer Aded Draftng, Desgn and Manufacturng Volume 5, Number, June 015, Page 14 CADDM Face Recognton Algorthm Fusng Monogenc Bnary Codng and Collaboratve Representaton FU Yu-xan, PENG Lang-yu College
More informationData 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 informationSupport 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 informationAn Entropy-Based Approach to Integrated Information Needs Assessment
Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology
More informationClassic Term Weighting Technique for Mining Web Content Outliers
Internatonal Conference on Computatonal Technques and Artfcal Intellgence (ICCTAI'2012) Penang, Malaysa Classc Term Weghtng Technque for Mnng Web Content Outlers W.R. Wan Zulkfel, N. Mustapha, and A. Mustapha
More informationOptimizing 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 informationA Weighted Method to Improve the Centroid-based Classifier
016 Internatonal onference on Electrcal Engneerng and utomaton (IEE 016) ISN: 978-1-60595-407-3 Weghted ethod to Improve the entrod-based lassfer huan LIU, Wen-yong WNG *, Guang-hu TU, Nan-nan LIU and
More informationSubspace 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 informationArabic Text Classification Using N-Gram Frequency Statistics A Comparative Study
Arabc Text Classfcaton Usng N-Gram Frequency Statstcs A Comparatve Study Lala Khresat Dept. of Computer Scence, Math and Physcs Farlegh Dcknson Unversty 285 Madson Ave, Madson NJ 07940 Khresat@fdu.edu
More informationFace 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 informationA Method of Hot Topic Detection in Blogs Using N-gram Model
84 JOURNAL OF SOFTWARE, VOL. 8, NO., JANUARY 203 A Method of Hot Topc Detecton n Blogs Usng N-gram Model Xaodong Wang College of Computer and Informaton Technology, Henan Normal Unversty, Xnxang, Chna
More informationA NOTE ON FUZZY CLOSURE OF A FUZZY SET
(JPMNT) Journal of Process Management New Technologes, Internatonal A NOTE ON FUZZY CLOSURE OF A FUZZY SET Bhmraj Basumatary Department of Mathematcal Scences, Bodoland Unversty, Kokrajhar, Assam, Inda,
More informationEfficient Text Classification by Weighted Proximal SVM *
Effcent ext Classfcaton by Weghted Proxmal SVM * Dong Zhuang 1, Benyu Zhang, Qang Yang 3, Jun Yan 4, Zheng Chen, Yng Chen 1 1 Computer Scence and Engneerng, Bejng Insttute of echnology, Bejng 100081, Chna
More informationIncremental Learnng wth Feature Shft Detecton for Personalzed E-mal Spam Flterng Gop Sanghan 1, Dr. Ketan Kotecha 2 1 Computer Engneerng Department, Nrma Unversty, Ahmedabad-382481, Inda. 2 Parul Unversty,
More informationWeb Document Classification Based on Fuzzy Association
Web Document Classfcaton Based on Fuzzy Assocaton Choochart Haruechayasa, Me-Lng Shyu Department of Electrcal and Computer Engneerng Unversty of Mam Coral Gables, FL 33124, USA charuech@mam.edu, shyu@mam.edu
More informationEdge 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 informationIncremental 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 informationOutline. 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 informationFuzzy Rough Neural Network and Its Application to Feature Selection
70 Internatonal Journal of Fuzzy Systems, Vol. 3, No. 4, December 0 Fuzzy Rough Neural Network and Its Applcaton to Feature Selecton Junyang Zhao and Zhl Zhang Abstract For the sake of measurng fuzzy uncertanty
More informationMulticlass Object Recognition based on Texture Linear Genetic Programming
Multclass Object Recognton based on Texture Lnear Genetc Programmng Gustavo Olague 1, Eva Romero 1 Leonardo Trujllo 1, and Br Bhanu 2 1 CICESE, Km. 107 carretera Tjuana-Ensenada, Mexco, olague@ccese.mx,
More informationEnhancement of Infrequent Purchased Product Recommendation Using Data Mining Techniques
Enhancement of Infrequent Purchased Product Recommendaton Usng Data Mnng Technques Noraswalza Abdullah, Yue Xu, Shlomo Geva, and Mark Loo Dscplne of Computer Scence Faculty of Scence and Technology Queensland
More informationFuzzy Logic Based RS Image Classification Using Maximum Likelihood and Mahalanobis Distance Classifiers
Research Artcle Internatonal Journal of Current Engneerng and Technology ISSN 77-46 3 INPRESSCO. All Rghts Reserved. Avalable at http://npressco.com/category/jcet Fuzzy Logc Based RS Image Usng Maxmum
More informationOn Evaluating Open Biometric Identification Systems
Proceedngs of Student/Faculty Research Day, CSIS, Pace Unversty, May 6th, 2005 On Evaluatng Open Bometrc Identfcaton Systems Mchael Gbbons, Sungsoo Yoon, Sung-Hyuk Cha and Charles Tappert mkegbb@us.bm.com,
More informationEYE 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 informationA 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 informationEnhanced Watermarking Technique for Color Images using Visual Cryptography
Informaton Assurance and Securty Letters 1 (2010) 024-028 Enhanced Watermarkng Technque for Color Images usng Vsual Cryptography Enas F. Al rawashdeh 1, Rawan I.Zaghloul 2 1 Balqa Appled Unversty, MIS
More informationAn 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 informationCollaboratively 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 informationSpam 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 informationClassification / 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 informationAn Accurate Evaluation of Integrals in Convex and Non convex Polygonal Domain by Twelve Node Quadrilateral Finite Element Method
Internatonal Journal of Computatonal and Appled Mathematcs. ISSN 89-4966 Volume, Number (07), pp. 33-4 Research Inda Publcatons http://www.rpublcaton.com An Accurate Evaluaton of Integrals n Convex and
More informationUnder-Sampling Approaches for Improving Prediction of the Minority Class in an Imbalanced Dataset
Under-Samplng Approaches for Improvng Predcton of the Mnorty Class n an Imbalanced Dataset Show-Jane Yen and Yue-Sh Lee Department of Computer Scence and Informaton Engneerng, Mng Chuan Unversty 5 The-Mng
More informationCHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION
48 CHAPTER 3 SEQUENTIAL MINIMAL OPTIMIZATION TRAINED SUPPORT VECTOR CLASSIFIER FOR CANCER PREDICTION 3.1 INTRODUCTION The raw mcroarray data s bascally an mage wth dfferent colors ndcatng hybrdzaton (Xue
More informationA 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 informationSkew 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 informationAudio Content Classification Method Research Based on Two-step Strategy
(IJACSA) Internatonal Journal of Advanced Computer Scence and Applcatons, Audo Content Classfcaton Method Research Based on Two-step Strategy Sume Lang Department of Computer Scence and Technology Chongqng
More informationThree 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 informationTraining of Kernel Fuzzy Classifiers by Dynamic Cluster Generation
Tranng of Kernel Fuzzy Classfers by Dynamc Cluster Generaton Shgeo Abe Graduate School of Scence and Technology Kobe Unversty Nada, Kobe, Japan abe@eedept.kobe-u.ac.jp Abstract We dscuss kernel fuzzy classfers
More informationCredibility Adjusted Term Frequency: A Supervised Term Weighting Scheme for Sentiment Analysis and Text Classification
Credblty Adjusted Term Frequency: A Supervsed Term Weghtng Scheme for Sentment Analyss and Text Classfcaton Yoon Km New York Unversty yhk255@nyu.edu Owen Zhang zhonghua.zhang2006@gmal.com Abstract We provde
More informationAssociative Based Classification Algorithm For Diabetes Disease Prediction
Internatonal Journal of Engneerng Trends and Technology (IJETT) Volume-41 Number-3 - November 016 Assocatve Based Classfcaton Algorthm For Dabetes Dsease Predcton 1 N. Gnana Deepka, Y.surekha, 3 G.Laltha
More informationKeyword-based Document Clustering
Keyword-based ocument lusterng Seung-Shk Kang School of omputer Scence Kookmn Unversty & AIrc hungnung-dong Songbuk-gu Seoul 36-72 Korea sskang@kookmn.ac.kr Abstract ocument clusterng s an aggregaton of
More information