A Misclassification Reduction Approach for Automatic Call Routing

Size: px
Start display at page:

Download "A Misclassification Reduction Approach for Automatic Call Routing"

Transcription

1 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. {fuceda, vllasen, mmontesg}@naoep.mx 2 Nuance Technologes, Mexco. Alejandro.Barbosa@nuance.com Abstract. Automatc call routng s one of the most mportant ssues n the call center doman. It can be modeled once performed the speech recognton of utterances as a text classfcaton task. Nevertheless, n ths case, texts are extremely small (just a few words) and there are a great number of narrow call-type classes. In ths paper, we propose a text classfcaton method specally suted to work on ths scenaro. Ths method consders a new weghtng scheme of terms and uses a multple stage classfcaton approach wth the am of balance the rate of rejected calls (drected to a human operator) and the classfcaton accuracy. The proposed method was evaluated on a Spansh corpus consstng of 24,638 call utterances achevng outstandng results: 95.5% of classfcaton accuracy wth a rejecton rate of just 8.2%. 1 Introducton Call routng (CR) s one of the most mportant ssues n the call center doman. It can be defned as the process of assocatng user requests wth ther desre destnatons [1]. In other words, t nvolves forwardng ncomng phone calls to someone qualfed to handle them (e.g., to the most approprate customer representatve). Tradtonally, ths task has been performed usng a touch-tone approach, whch allows users to navgate through dfferent optons by a herarchcal menu. In spte of ts conceptual smplcty, ths approach has several dsadvantages, such as the tme the users need for lstenng all optons as well as the dffculty for matchng exstng optons wth ther specfc needs. These nconvenences are evdent even for smple requests such as I want to know my account balance, whch may requre users to navgate as many as four or fve nested menus wth four or fve optons each [2]. The nconvenences of the tone-touch approach leaded to the emergence of automatc call routng systems, whch allow users to nteract n natural spoken language [1], and therefore, help companes to sgnfcantly reduce ther telephone tme costs. Typcally, ths knd of systems consders two man steps. In the frst one, phone calls are transcrbed to text by means of a speech recognton process. Then, n a second step, text transcrptons are used to predct the correct destnaton for the phone calls.

2 In partcular, ths paper focuses on the second step, whch, n fact, can be consdered as a text classfcaton problem. Text classfcaton, the assgnment of free text documents to one or more predefned categores based on ther content, s a wdely studed problem that has recently acheved sgnfcant advances [3]. Nevertheless, t s mportant to pont out that automatc call routng s a more dffcult problem than ordnary text classfcaton, snce phone call transcrptons are very short and nosy texts. That s, they are texts consstng of just a few words and contanng many errors such as word nsertons, deletons and substtutons. In addton, automatc call routng dffers from conventonal text classfcaton n that, n the former, there s always the possblty of appealng for human assstance. In partcular, call routng systems tend to reject all ambguous or ncomprehensble transcrptons, transferrng ther resoluton to a human operator. Evdently, ths crcumstance s of great relevance, snce havng a hgh rejecton rate mples a hgh telephone tme cost, whereas havng a low classfcaton accuracy (due to the msclassfcaton of unclear transcrptons) causes a strong dscomfort among users. Therefore, one of the man research topcs on automatc call routng s to provde a good balance between these two factors, the classfcaton accuracy and the rejecton rate. In ths paper, we propose a classfcaton method specally suted to the task of automatc call routng. On the one hand, ths method consders a new weghtng scheme that s especally approprate for short texts. On the other hand, t uses a twostep classfcaton approach that allows achevng an adequate balance between the rejecton rate and the classfcaton accuracy. In a frst step, ths approach guarantees hgh classfcaton accuracy by toleratng a hgh rejecton rate. Subsequently, n a second step, t reclassfy rejected transcrptons wth the ntenton of recoverng some nstances and, therefore, reducng the fnal rejecton rate. Expermental results on a Spansh corpus consstng of 24,638 call utterances are encouragng; the proposed method could correctly classfed 95.5% of the utterances wth a rejecton rate of just 8.2%, outperformng the applcaton of other text classfcaton approaches. The rest of the paper s organzed as follows. Secton 2 descrbes some prevous work on automatc call routng. Secton 3 descrbes the proposed method for automatc call routng. Secton 4 presents the evaluaton results. Fnally, Secton 5 gves our conclusons and descrbes some deas for future work. 2 Related work There exst dfferent approaches for automatc call routng; among them we can menton the followng three ones: 1. A straghtforward approach based on the use of keywords for trggerng dfferent destnatons [4]. 2. An approach based on the use of language models that descrbe the man characterstcs (words sequences) of each destnaton [5]. 3. An approach based on the applcaton of tradtonal text classfcaton methods [2,6].

3 From these approaches, the frst one s the smplest to understand and mplement; nevertheless, t s very sensble to transcrpton errors. The second approach s more robust than the former, but t requres more tranng data as well as clearly dfferentable destnaton vocabulares. Fnally, the thrd approach s less sensble to transcrpton errors, but t s damaged by the short length of transcrptons as well as by the great number of narrow destnatons (categores). Dfferent works have evaluated the effectveness of conventonal text classfcaton methods for automatc call routng. For nstance, [7] shows a comparson of several learnng algorthms on a corpus of 4000 manual transcrptons correspondng to nne balanced categores. It reports a result of 81.6% of accuracy usng a smple recurrent network. In the same way, [8] reports an accuracy of 75.9%; nevertheless, n ths case they used a small corpus of 743 automatc transcrptons from sx dfferent categores. Both works do not report the rejecton rate. More recently, some works have proposed dfferent adaptatons to the tradtonal text classfcaton approach. These adaptatons consder the use of dfferent utterance characterzatons, weghtng schemes, feature selecton methods and learnng strateges. For nstance, [9] presents a learnng approach that uses a cascade of bnary classfers (of Support Vector Machnes) wth the am of reducng the classfcaton errors. In ths approach, those nstances that could not be classfed (.e., that pass through all classfers) are rejected. Reported results on a set of automatc transcrptons from 19 categores are very relevant; on the one hand, t reports a best accuracy of 99% correspondng to a rejecton rate of 75%, and on the other hand, a best rejecton rate of 15% for a classfcaton accuracy of 95%. Smlar to these works, our automatc call routng method s also based on a supervsed learnng approach. Its man dfference relays on the use of a novel two-step classfcaton approach that allows mantanng a good balance between the accuracy and rejecton rates. In addton, t apples a weghtng scheme that s less sensble to the short length of texts. The followng secton ntroduces the proposed method. 3 Proposed Method As we prevously mentoned, our method ncludes two major modfcatons to the tradtonal text classfcaton approach. On the one hand, t uses a new term weghtng scheme that allows mprovng the dscrmnaton among narrow classes formed by very short texts. On the other hand, t consders a two-step classfcaton approach that helps mantanng an adequate balance between the accuracy and rejecton rates. In the followng subsectons, we descrbe n detal these two modfcatons. 3.1 Term Weghtng Scheme Typcally, term relevance s determned usng the well-known tf-df weghtng scheme. Ths scheme, however, presents some drawbacks for ts applcaton to automatc call routng. Frst, gven the short length of texts, the tf-value s qute smlar

4 for most of the terms from a category 1. Second, gven the presence of several related narrow categores, the df-value s also very smlar for most terms. Therefore, ths weghtng scheme does not allow achevng a good dscrmnaton among categores. Based on the prevous observatons, we propose a weghtng scheme that computes the weght of terms n relaton to ther occurrence n the whole categores nstead than n ndvdual utterances. Ths scheme consders that the probablty of occurrence of a term across dstnct categores s dfferent. In partcular, we compute the weght of a term t n a category c C as follows: c c p t wt = c (1) max p p c t k c = P( c c ( ) k t) = c t ck ft c C where C s the set of categores, f t s the frequency of occurrence of the term t n c the category c, and pt ndcates the condtonal probablty P(c t), that s, the probablty of havng the category c gven the presence of the term t. It s nterestng to notce that the fnal weght of a term n a category s obtaned by applyng a knd of normalzaton wth respect to the probabltes of other terms from the same category. Roughly speakng, ths normalzaton allows gvng greater weght to terms clearly related to the category and reducng the value of terms unformly dstrbuted among several classes. 3.2 Two-Step Classfcaton Approach Fgure 1 shows the general archtecture of the proposed classfcaton approach, whch conssts of two man steps: a hgh-precson classfcaton step and a msclassfcaton reducton step. The man purpose of ths new approach s to acheve a good balance between the classfcaton accuracy and the rejecton rate. The frst step acheves an ntal classfcaton of the nput utterances (.e., the phone call transcrptons). Ths classfcaton consders all categores ncludng an unknown category that gathers all ambguous and ncomprehensble transcrptons. Ths category functons as a rejecton class, snce ts elements are transferred to a human operator for ther resoluton. Ths way, the purpose of ths step s to guarantee a hgh classfcaton accuracy by toleratng a hgh rejecton rate. In other words, n ths step only the most confdent utterances are sent to content categores, whereas the rest of them are classfed as unknown. Subsequently, n a second step, the dea s to reclassfy the set of rejected transcrptons wth the ntenton of recoverng some utterances and, therefore, reducng the fnal rejecton rate. Ths step combnes two addtonal classfers. One of them consders only two categores (unknown and the rest), and t s especally suted to dstngush clear ambguous and ncomprehensble transcrptons. The other one k f 1 Ths s because users tend to use a smple and drect language (more restrcted vocabulary) when they nteract wth a machne (an automatc system).

5 Hgh precson classfcaton step Input utterance Frst Classfer (n classes) Utterance classfed as Unknown NO Category of utterance YES Msclassfcaton reducton step Second Classfer (2 classes) Utterance classfed as Unknown NO Thrd Classfer (n 1 classes) YES Rejected utterance Fgure 1. General archtecture of the proposed method consders all categores except unknown and ts purpose s to reallocate rescue utterances wthn content categores. It s mportant to pont out that the weght of terms (defned n accordance to formula 1) vares from one classfer to another because they consder dfferent sets of categores. Ths characterstc of our approach s very relevant, snce t allows gvng the terms a convenent (hghly dscrmnatve) weght at each step. 4 Expermental Evaluaton 4.1 Tran and Test Corpora In order to evaluate the proposed method we used a corpus consstng of 24,638 Spansh automatc transcrptons 2. These transcrptons came from a company that provdes TV, nternet and telephone servces, and were manually classfed n 24 dfferent categores. The frst 23 categores correspond to content categores (.e., to possble destnatons for user requests), whereas the last one s an unknown category, whch gathers all utterances that could not be classfed and that must be rejected. For evaluaton purposes, we dvded the corpus n two non-overlapped subsets: a tran set formed by 80% of the nstances from each category (n total, 19,694 utterances), and a test set ncludng the remanng 20% of the nstances (4,944 utterances). Fgure 2 shows the dstrbuton of nstances per category; as t can be observed, ths s a very mbalanced corpus. 2 It was assembled by Nuance technologes ( usng the Wzard of Oz technque.

6 PaymentPlan TVOnDemand PhoneSupport PhoneChange VagueInternet ServceStart TelephoneStart SupportAtHome CableSuppo rt AccountB alance VagueCable CableStart ServceChange AccountInfo AddressChange TechSuport InternetSupport CableChange InternetStart Payments Retentons Representatve VaguePhone Unkno wn 0.00% 5.00% 10.00% 15.00% 20.00% 25.00% Fgure 2. Percentage of nstances per category n the used corpus We decded usng ths evaluaton scheme because we wshed to smulate a real stuaton, where t s possble to fnd utterances contanng unspecfed words for the classfer. In partcular, the vocabulares from the tran and test sets showed mportant dfferences: 13.5% of the words from the test nstances were not ncluded n the tran set. 4.2 Results In order to obtan a general vson from the proposed method we performed several experments. Intally, we studed the usefulness of the weghtng scheme for tunng the accuracy and rejecton rates. Then, we evaluated the effectveness of our twostep classfcaton approach by comparng ts results aganst others from conventonal one-step classfers. Tunng the accuracy and rejecton rates For ths frst experment, we used the weght of terms as the man crteron for feature selecton. In partcular, we elmnated terms wth weght n all categores less than a gven specfed threshold. Table 1 presents the results correspondng to dfferent thresholds. It manly shows the number of tranng features, the classfcaton accuracy 3 as well as the rejecton rate 4. It s mportant to menton that n all cases we 3 Defned as the percentage of utterances correctly classfed wthn the set of 23 content categores. 4 Defned as the percentage of test utterances classfed as unknown.

7 used a sngle (one-step) mult-class Naïve Bayes classfer and appled a Boolean weghtng for the test utterances 5. Table 1. Feature selecton usng the proposed weghtng scheme Weght threshold Number of features Classfcaton accuracy Rejecton rate % 1.0% % 0.9% % 24.2% % 31.5% % 40.7% % 50.9% Results from ths ntal experment showed that usng dfferent thresholds t was possble to obtan dfferent combnatons of classfcaton accuraces and rejecton rates. As we expected, the use of the most dscrmnatve terms allowed achevng better classfcaton accuracy, but also ncremented the rejecton rate snce several nstances could not be represented. These results are of great relevance to our method snce t consders three dfferent classfers havng dfferent purposes. Evaluatng the effectveness of our two-step classfcaton method The goal of the proposed two-step classfcaton method s to acheve a good balance between the classfcaton accuracy and the rejecton rate. In order to evaluate the effectveness of ths method, we compared ts results aganst those from a conventonal one-step classfer (refer to Table 1). Our two-step classfcaton method, as shown n Fgure 1, consders the combnaton of three dfferent classfers. All of them, as n the prevous experment, were mplemented usng a Naïve Bayes classfer and Boolean weghts for representng test nstances. Nevertheless, n ths case, each classfer consders a dfferent weght threshold 6 : the frst two classfers used a threshold of 0.3, whereas the last one used a threshold of 0.2. Table 2. Results from the proposed two-step classfcaton method Correctly classfed Utterances Frst step Second step Classfcaton accuracy Rejecton rate % 8.2% Table 2 shows the results acheved by ths method. In ths case, the accuracy was calculated takng nto consderaton the correctly classfed nstances from the frst step as well as the correctly rescued nstances from the second step. On the other hand, the rejecton rate was obtaned from the set of nstances classfed as unknown 5 In contrast, tranng nstances were represented usng the proposed weghtng scheme. 6 These thresholds were emprcally determned; for each classfer we evaluated several thresholds and selected the combnaton that acheved the best overall accuracy.

8 by the second classfer. As can be observed, the acheved results are encouragng; the proposed method reached an adequate balance between the classfcaton accuracy and rejecton rate, better than the results from all prevous one-step classfers (refer to Table 1). Fnally, n order to have more nformaton to judge the effectveness of the proposed method, we compared ts results aganst those from the AdaBoost algorthm (commonly used n ths task [10]). Table 3 shows the results from ths comparson. The frst row presents the results of our method, and the second and the thrd rows show the results from AdaBoost usng Naïve Bayes as base classfer and applyng sx teratons. The only dfference from these two runs was the used weghtng scheme. The second row reports the result usng the proposed weghtng scheme, whereas the last row shows the result correspondng to the tf-df weghtng scheme. Table 3. The proposed method aganst the AdaBoost classfer Classfcaton scheme Classfcaton accuracy Rejecton rate Recall of unknown class Our method 95.5% 8.2% 100% AdaBoost (usng our weghtng scheme) 98.9% 29.7% 100% AdaBoost (usng tf-df weghts) 78.2% 6.8% 60% The results from ths experment are very nterestng; they show that our method allowed achevng a better balance between the accuracy and rejecton rates than the tradtonal AdaBoost method. On the other hand, they also show that the proposed weghtng scheme lead to better results than the tf-df weghtng scheme. Moreover, usng the tf-df weghtng scheme t was possble to obtan a small rejecton rate, however, t could not dentfy (and therefore, transfer to a human operator) all ambguous or ncomprehensble utterances. Ths last factor s very relevant snce the msclassfcaton of unclear transcrptons tend to cause a strong dscomfort among users. 5 Conclusons Ths paper proposed a new text classfcaton method that s specally suted to the task of automatc call routng. Ths method dffers from prevous works n two man concerns: the applcaton of a new term weghtng scheme, and the use of a two-step classfcaton approach. Expermental results on a Spansh corpus consstng of 24,638 call utterances showed, on the one hand, that the proposed weghtng scheme s especally approprate for short texts, and on the other hand, that the two-step classfcaton approach allows mantanng an adequate balance between the classfcaton accuracy and the rejecton rate. In other words, these results ndcated that the proposed method s pertnent for the automatc call routng task, whch commonly consders a great number of narrow categores, mbalanced tranng data sets as well as very short text nstances.

9 Another mportant aspect s that the method tself does not rely on language dependent rules, and therefore, t would be easly used n other languages. Fnally, t s mportant to menton that the method can be adjusted (movng the weght thresholds) for each partcular stuaton, n order to reduce the rejecton rate or mprove the classfcaton accuracy. In addton, t s also mportant to comment that the performance of ths method greatly depends on the corpus sze. It requres of a large tran corpora to calculate confdent weghts for the terms. For ths reason, one of the man drectons for future work s the applcaton of sem-supervsed learnng strateges. Acknowledgements: Ths work was done under partal support of CONACYT project grant References 1. Lee, C.-H., Carpenter, B., Chou, W., Chu-Carroll, J., Rechl, W., Saad, A. & Zhou, Q. On Natural Languge Call Routng. Speech Communcaton, Vol. 31, Issue 4, August 2000, pp Carpenter B. & Chu-Carroll, J. Natural Language Call Routng: A Robust, Self-organzng Approach, Proc. ICSLP Sebastan F., Machne learnng n automated text categorzaton, ACM Computng Surveys, 34(1):1 47, Gorn, A.L., Rccard, G. & Wrght, J.H. How may I help you? Speech Communcaton 23, Elsever. 5. Huang Q. & Cox, S.J. Automatc Call Routng wth Multple Language Models. Proc. Workshop on Spoken Language Understandng for Conversatonal Systems, Boston, May Kuo, H.-K., J., Lee, C.-H., Ztoun, I., Fosler-Lusser, E. & Ammcht, E. Dscrmnatve tranng for call classfcaton and routng. In ICSLP-2002, Garfeld S., Wermter S. & Devln S., Spoken Language Classfcaton usng Hybrd Classfer Combnaton. Internatonal Journal of Hybrd Intellgent Systems, Vol. 2, Issue 1, pp 13 33, January Mng Tang, Bryan Pellom, Kadr Hacoglu. Call-Type Classfcaton And Unsupervsed Tranng For The Call Center Doman. Automatc Speech Recognton and Understandng IEEE Workshop, , ISBN ASRU Haffner, P., Tur, G., Wrght, J.H. Optmzng SVMs for complex call classfcaton. Internatonal Conference on Acoustcs, Speech, and Sgnal Processng. ICASSP '03 IEEE. Aprl 2003 Vol. 1, pp , Schapre, R. E. & Snger, Y. BoosTexter: A Boostng-based System for Text Categorzaton. Machne Learnng, 39, pp , Kluwer.

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

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

Classifier Selection Based on Data Complexity Measures *

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

More information

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

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

More information

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

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

Optimizing Document Scoring for Query Retrieval

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

More information

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

User Authentication Based On Behavioral Mouse Dynamics Biometrics

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

More information

A 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

Classifying Acoustic Transient Signals Using Artificial Intelligence

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

More information

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

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

Complex Numbers. Now we also saw that if a and b were both positive then ab = a b. For a second let s forget that restriction and do the following.

Complex Numbers. Now we also saw that if a and b were both positive then ab = a b. For a second let s forget that restriction and do the following. Complex Numbers The last topc n ths secton s not really related to most of what we ve done n ths chapter, although t s somewhat related to the radcals secton as we wll see. We also won t need the materal

More information

UB at GeoCLEF Department of Geography Abstract

UB 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 information

A Fast Content-Based Multimedia Retrieval Technique Using Compressed Data

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

More information

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

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

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

More information

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

Face Detection with Deep Learning

Face Detection with Deep Learning Face Detecton wth Deep Learnng Yu Shen Yus122@ucsd.edu A13227146 Kuan-We Chen kuc010@ucsd.edu A99045121 Yzhou Hao y3hao@ucsd.edu A98017773 Mn Hsuan Wu mhwu@ucsd.edu A92424998 Abstract The project here

More information

Incremental Learning with Support Vector Machines and Fuzzy Set Theory

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

More information

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

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

More information

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

Experiments in Text Categorization Using Term Selection by Distance to Transition Point

Experiments 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 information

Performance Evaluation of Information Retrieval Systems

Performance 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 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

A Novel Term_Class Relevance Measure for Text Categorization

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

More information

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

Gender Classification using Interlaced Derivative Patterns

Gender 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 information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

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

More information

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

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

More information

Cluster Analysis of Electrical Behavior

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

More information

Analysis of Continuous Beams in General

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

More information

Load-Balanced Anycast Routing

Load-Balanced Anycast Routing Load-Balanced Anycast Routng Chng-Yu Ln, Jung-Hua Lo, and Sy-Yen Kuo Department of Electrcal Engneerng atonal Tawan Unversty, Tape, Tawan sykuo@cc.ee.ntu.edu.tw Abstract For fault-tolerance and load-balance

More information

BAYESIAN MULTI-SOURCE DOMAIN ADAPTATION

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

More information

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

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

More information

Benchmarking of Update Learning Strategies on Digit Classifier Systems

Benchmarking of Update Learning Strategies on Digit Classifier Systems 2012 Internatonal Conference on Fronters n Handwrtng Recognton Benchmarkng of Update Learnng Strateges on Dgt Classfer Systems D. Barbuzz, D. Impedovo, G. Prlo Dpartmento d Informatca Unverstà degl Stud

More information

Reducing Frame Rate for Object Tracking

Reducing Frame Rate for Object Tracking Reducng Frame Rate for Object Trackng Pavel Korshunov 1 and We Tsang Oo 2 1 Natonal Unversty of Sngapore, Sngapore 11977, pavelkor@comp.nus.edu.sg 2 Natonal Unversty of Sngapore, Sngapore 11977, oowt@comp.nus.edu.sg

More information

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

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

More information

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

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

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution

Real-time Motion Capture System Using One Video Camera Based on Color and Edge Distribution Real-tme Moton Capture System Usng One Vdeo Camera Based on Color and Edge Dstrbuton YOSHIAKI AKAZAWA, YOSHIHIRO OKADA, AND KOICHI NIIJIMA Graduate School of Informaton Scence and Electrcal Engneerng,

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

Deep Classification in Large-scale Text Hierarchies

Deep 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 information

Online Detection and Classification of Moving Objects Using Progressively Improving Detectors

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

More information

The Codesign Challenge

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

More information

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

Available online at Available online at Advanced in Control Engineering and Information Science

Available 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 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

D. Barbuzzi* and G. Pirlo

D. Barbuzzi* and G. Pirlo Int. J. Sgnal and Imagng Systems Engneerng, Vol. 7, No. 4, 2014 245 bout retranng rule n mult-expert ntellgent system for sem-supervsed learnng usng SVM classfers D. Barbuzz* and G. Prlo Department of

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

Audio Content Classification Method Research Based on Two-step Strategy

Audio 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 information

SVM-based Learning for Multiple Model Estimation

SVM-based Learning for Multiple Model Estimation SVM-based Learnng for Multple Model Estmaton Vladmr Cherkassky and Yunqan Ma Department of Electrcal and Computer Engneerng Unversty of Mnnesota Mnneapols, MN 55455 {cherkass,myq}@ece.umn.edu Abstract:

More information

Machine Learning: Algorithms and Applications

Machine 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 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

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

An Entropy-Based Approach to Integrated Information Needs Assessment

An 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 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

BOOSTING CLASSIFICATION ACCURACY WITH SAMPLES CHOSEN FROM A VALIDATION SET

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

More information

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

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

More information

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity

Corner-Based Image Alignment using Pyramid Structure with Gradient Vector Similarity Journal of Sgnal and Informaton Processng, 013, 4, 114-119 do:10.436/jsp.013.43b00 Publshed Onlne August 013 (http://www.scrp.org/journal/jsp) Corner-Based Image Algnment usng Pyramd Structure wth Gradent

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

CAN COMPUTERS LEARN FASTER? Seyda Ertekin Computer Science & Engineering The Pennsylvania State University

CAN 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 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

Data Mining: Model Evaluation

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

More information

Parallel matrix-vector multiplication

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

More information

Connection-information-based connection rerouting for connection-oriented mobile communication networks

Connection-information-based connection rerouting for connection-oriented mobile communication networks Dstrb. Syst. Engng 5 (1998) 47 65. Prnted n the UK PII: S0967-1846(98)90513-7 Connecton-nformaton-based connecton reroutng for connecton-orented moble communcaton networks Mnho Song, Yanghee Cho and Chongsang

More information

Decision Strategies for Rating Objects in Knowledge-Shared Research Networks

Decision Strategies for Rating Objects in Knowledge-Shared Research Networks Decson Strateges for Ratng Objects n Knowledge-Shared Research etwors ALEXADRA GRACHAROVA *, HAS-JOACHM ER **, HASSA OUR ELD ** OM SUUROE ***, HARR ARAKSE *** * nsttute of Control and System Research,

More information

Feature Selection for Natural Language Call Routing Based on Self-Adaptive Genetic Algorithm

Feature 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 information

Federated Search of Text-Based Digital Libraries in Hierarchical Peer-to-Peer Networks

Federated Search of Text-Based Digital Libraries in Hierarchical Peer-to-Peer Networks Federated Search of Text-Based Dgtal Lbrares n Herarchcal Peer-to-Peer Networks Je Lu School of Computer Scence Carnege Mellon Unversty Pttsburgh, PA 15213 jelu@cs.cmu.edu Jame Callan School of Computer

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

A Background Subtraction for a Vision-based User Interface *

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

More information

Investigating the Performance of Naïve- Bayes Classifiers and K- Nearest Neighbor Classifiers

Investigating 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 information

Feature Kernel Functions: Improving SVMs Using High-level Knowledge

Feature Kernel Functions: Improving SVMs Using High-level Knowledge Feature Kernel Functons: Improvng SVMs Usng Hgh-level Knowledge Qang Sun, Gerald DeJong Department of Computer Scence, Unversty of Illnos at Urbana-Champagn qangsun@uuc.edu, dejong@cs.uuc.edu Abstract

More information

Simulation Based Analysis of FAST TCP using OMNET++

Simulation Based Analysis of FAST TCP using OMNET++ Smulaton Based Analyss of FAST TCP usng OMNET++ Umar ul Hassan 04030038@lums.edu.pk Md Term Report CS678 Topcs n Internet Research Sprng, 2006 Introducton Internet traffc s doublng roughly every 3 months

More information

GA-Based Learning Algorithms to Identify Fuzzy Rules for Fuzzy Neural Networks

GA-Based Learning Algorithms to Identify Fuzzy Rules for Fuzzy Neural Networks Seventh Internatonal Conference on Intellgent Systems Desgn and Applcatons GA-Based Learnng Algorthms to Identfy Fuzzy Rules for Fuzzy Neural Networks K Almejall, K Dahal, Member IEEE, and A Hossan, Member

More information

DESIGNING EFFICIENT MULTIMODAL CLASSIFICATION SYSTEMS BASED ON FEATURES AND SVM KERNELS SELECTION

DESIGNING EFFICIENT MULTIMODAL CLASSIFICATION SYSTEMS BASED ON FEATURES AND SVM KERNELS SELECTION DESIGNING EFFICIENT MULTIMODAL CLASSIFICATION SYSTEMS BASED ON FEATURES AND SVM KERNELS SELECTION Anca APATEAN Techncal Unversty of Cluj-Napoca, Communcatons Department Address, Phone and Fax Numbers,

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

Under-Sampling Approaches for Improving Prediction of the Minority Class in an Imbalanced Dataset

Under-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 information

Heterogeneous Visual Features Fusion via Sparse Multimodal Machine

Heterogeneous Visual Features Fusion via Sparse Multimodal Machine 013 IEEE Conference on Computer Vson and Pattern Recognton Heterogeneous Vsual Features Fuson va Sparse Multmodal Machne Hua ang, Fepng Ne, Heng Huang, Chrs Dng Department of Electrcal Engneerng and Computer

More information

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

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

More information

An Efficient Face Detection Method Using Adaboost and Facial Parts

An Efficient Face Detection Method Using Adaboost and Facial Parts An Effcent Face Detecton Method Usng Adaboost and Facal Parts Yasaman Heydarzadeh, Abolfazl Torogh Haghghat Computer, IT and Electronc department Azad Unversty of Qazvn Tehran, Iran heydarzadeh@ qau.ac.r,

More information

A MODIFIED K-NEAREST NEIGHBOR CLASSIFIER TO DEAL WITH UNBALANCED CLASSES

A 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 information

Face Recognition Based on SVM and 2DPCA

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

More information

An Improved Neural Network Algorithm for Classifying the Transmission Line Faults

An Improved Neural Network Algorithm for Classifying the Transmission Line Faults 1 An Improved Neural Network Algorthm for Classfyng the Transmsson Lne Faults S. Vaslc, Student Member, IEEE, M. Kezunovc, Fellow, IEEE Abstract--Ths study ntroduces a new concept of artfcal ntellgence

More information

Fast Feature Value Searching for Face Detection

Fast Feature Value Searching for Face Detection Vol., No. 2 Computer and Informaton Scence Fast Feature Value Searchng for Face Detecton Yunyang Yan Department of Computer Engneerng Huayn Insttute of Technology Hua an 22300, Chna E-mal: areyyyke@63.com

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

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

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

More information

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

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

More information

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

Query Clustering Using a Hybrid Query Similarity Measure

Query 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 information

Backpropagation: In Search of Performance Parameters

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

More information

Pruning Training Corpus to Speedup Text Classification 1

Pruning 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 information

Problem Set 3 Solutions

Problem Set 3 Solutions Introducton to Algorthms October 4, 2002 Massachusetts Insttute of Technology 6046J/18410J Professors Erk Demane and Shaf Goldwasser Handout 14 Problem Set 3 Solutons (Exercses were not to be turned n,

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

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

Recognition of Handwritten Numerals Using a Combined Classifier with Hybrid Features

Recognition of Handwritten Numerals Using a Combined Classifier with Hybrid Features Recognton of Handwrtten Numerals Usng a Combned Classfer wth Hybrd Features Kyoung Mn Km 1,4, Joong Jo Park 2, Young G Song 3, In Cheol Km 1, and Chng Y. Suen 1 1 Centre for Pattern Recognton and Machne

More information

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices

Steps for Computing the Dissimilarity, Entropy, Herfindahl-Hirschman and. Accessibility (Gravity with Competition) Indices Steps for Computng the Dssmlarty, Entropy, Herfndahl-Hrschman and Accessblty (Gravty wth Competton) Indces I. Dssmlarty Index Measurement: The followng formula can be used to measure the evenness between

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