On Evaluating Open Biometric Identification Systems

Size: px
Start display at page:

Download "On Evaluating Open Biometric Identification Systems"

Transcription

1 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 {scha, Abstract Ths paper concerns the generalzablty of bometrc dentfcaton n open systems. Many researchers have clamed hgh dentfcaton accuraces on closed system consstng of a few hundred or thousand members. Here, we consder what happens to these closed dentfcaton systems as they are opened to non-members. We clam that these systems do not generalze well as the non-member populaton ncreases. To support ths clam, we frst take a look at the vsualzaton of pattern classfcaton usng Support Vector Machnes (SVM), Nearest Neghbor (NN) and Artfcal Neural Network (ANN). Next, we present expermental results on wrter and rs bometrc databases usng the afore mentoned classfers. We fnd that system securty (1-FAR) decreases rapdly for closed systems when they are tested n open-system mode as the number of non members tested ncreases. We also fnd that, although systems can be traned for greater closed-system securty usng SVM rather than NN classfers, the NN classfers are better for generalzng to open systems due to ther superor capablty of rejectng non members. Keywords Bometrc dentfcaton, Wrter dentfcaton, Irs dentfcaton, Nearest Neghbor, Support Vector Machnes 1 INTRODUCTION Bometrc applcatons are becomng more common and acceptable n today's socety. Technology contnues to mprove, provdng faster processors, smaller sensors and cheaper materals, all of whch are contrbutng to relable, affordable bometrc applcatons. The most common use of bometrcs s for verfcaton. In bometrc verfcaton systems, a user s dentfed by an ID or smart card and s verfed by ther bometrc,.e., a person s bologcal or behavoral characterstc such as ther fngerprnt, voce, rs, or sgnature. Ths s analogous to a user at an ATM machne usng a bank card to dentfy and a PIN to verfy. Another use of bometrcs s for dentfcaton, whch s the focus of ths paper. Identfcaton can be appled n a closed system such as employee postve dentfcaton for buldng access, or n an open system such as a natonal ID system or negatve dentfcaton for passenger screenng at an arport. Postve bometrc dentfcaton, a 1-to-many problem, s more challengng than verfcaton, a 1-to-1 problem. As stated n [1], "postve dentfcaton s perhaps the most ambtous use of bometrcs technology." There have been many promsng results reported for closed dentfcaton systems. Although hgh accuraces have been reported n wrter, rs, and hand geometry studes [4, 10, 12, 15, 17], these accuraces may lead to a false mpresson of securty. One may ask whether there really are any stuatons that correspond to closed worlds[1]. For example, take an employee dentfcaton system. Can t be guaranteed that a bometrc of a guest (a non-member) vstng the faclty does not match one of an employee (a member)? Ths paper wll nvestgate the generalzablty of bometrc dentfcaton as t pertans to the securty of an open system. Our hypothess s that the accuraces reported for closed systems are relevant only to those systems and do not generalze well to larger, open systems contanng non-members. Snce t s mpractcal to test a true populaton, we use a reverse approach to support the hypothess. We wll work wth a database M of m members, but assume a closed system of m ' members, where m' < m, and tran the system on the m ' members. We then have m m' members to test how well the system holds up when non-members attempt to enter the system. Ths approach s used on two bometrc databases, one consstng of wrter data and the other of rs data. In secton 2 of ths paper, postve dentfcaton and the assocated error rates wll be explaned. In secton 3, the bometrc databases and the pattern classfcaton technques used n ths paper wll be descrbed. In secton 4, the statstcal experments to support the hypothess are descrbed and observatons presented. Secton 5 concludes wth a summary and consderatons for future work. 2 ERROR RATE TYPES IN BIOMETRIC IDENTIFI- CATION Consder the postve dentfcaton model. Postve dentfcaton refers to determnng that a gven ndvdual s n a member database[1]. Ths s an m-class classfcaton problem that s, gven data from m subjects n a bometrc database M = { s1, s2, L, s m }, the problem s to dentfy an unknown bometrc sample d from a person,, where s q D5.1

2 s q M. In ths model, a classfer can be traned based on d all exemplars n M to fnd decson boundares, e.g., support vector machne. If a smlarty-based classfer such as a nearest neghbor s used, an unknown sample d s compared to each of M. The error rate n ths model s smply a number of msclassfed nstances dvded by the testng set sze. We clam that a classfer wth the lowest error rate s not necessarly the best for securty, and that classfer desgners mght consder the followng three types of error rates. In ths paper, we wll pay close attenton to the Securty measurement as we test our hypothess. Consder an unknown bometrc sample d from a person, sq, where sq M. If ths bometrc data of a non member enters drectly to the above model, can t be classfed correctly? If the classfer has no reject capablty, the unknown wll be classfed nto one of the decson regons or as the closest matchng subject. However, f the classfer has a reject capablty, the number of classes n M becomes m +1,.e., m member classes + 1 reject class. Therefore, f the questoned nstance s n a reject area n SVM or the closest match s outsde the nearest neghbor thresholds, the unknown wll be classfed as none of the members. Ths study nvestgates the reject capablty of two classfers: support vector machne and nearest neghbor. In the later scenaro wth members and non-members, there are three knds of error. A false reject, FR, error occurs when a classfer dentfes an unknown bometrc sample d from a person, sq, where sq M, as a reject. The other errors are false accepts, FA, of whch there are two types those that can occur between members of the system, FA (1), and those that can occur as non-members enter the system, FA (2). FA(1) occurs when a classfer dentfes an unknown bometrc sample d from a person, sq to s where sq, s M and sq s. FA(2) occurs when a classfer dentfes an unknown bometrc sample d from a person, sq to s, where sq M and s M. Fgure 1 llustrates these three error types. The frequences at whch the false accepts and false rejects occur are known as the False Accept Rate (FAR) and the False Reject Rate (FRR), respectvely. These two error rates are used to determne the two key performance measurements of a bometrc system: convenence and securty [1]: Convenence = 1 FRR Securty = 1 FAR (1) Fgure 1. (a) The graph dsplays classfcaton boundares for a hypothetcal two-member database usng the Nearest Neghbor classfer wth threshold t. False accepts and false rejects can occur between members of the system, and false accepts can also occur as non-members enter the system. (b) Same as (a) but wth a three-member database and SVM classfer. 3 BIOMETRIC DATABASES AND CLASSIFIERS Two bometrc databases are used to support our clams n ths study: the wrter and rs bometrc databases. Although there are many classfers to choose from n the feld of pattern classfcaton, we used two pattern classfcaton technques: Support Vector Machnes (SVM) and Nearest Neghbor. 3.1 Databases In a prevous study, Cha et al. [3] studed the ndvdualty of handwrtng usng a database of handwrtng samples D5.2

3 from 841 subjects representatve of the Unted States populaton. Each subject coped three tmes a source document contanng 156 words carefully constructed to have each letter of the alphabet used n the startng (both upper and lowercase), mddle (lowercase), and endng poston (lowercase) of a word. Each document was dgtzed and features were extracted at the document, word, and character level. For the purposes of ths study, we used the same database but focus only on the document features: entropy, threshold, number of black pxels, number of exteror contours, number of nteror contours, slant, heght, horzontal slope, vertcal slope, negatve slope, and postve slope. A detaled explanaton of these features can be found n [4]. From the rs bometrc mage database [9], we selected 10 left bare eye samples of 52 subjects. In comparson to the wrter database, the rs database has many fewer subjects, but a much larger number of samples per subject. Ths wll allow for more samples to be traned. After the mages are acqured, they are segmented to provde a normalzed rectangular sample of the rs. Features are extracted usng 2-D mult-level wavelet transforms. For ths experment, 3 levels are used producng a total of 12 parts. The 12 parts produce 12 feature vectors consstng of the coeffcents from the wavelet transform. The mean and varance of each vector are obtaned to produce a total of 24 features for each sample. See [16] for more nformaton on the 2-D wavelet transforms used. 3.2 Classfers In recent years, the SVM classfer has ganed consderable popularty among the possble classfers. The objectve of the SVM classfer s to separate data wth a maxmal margn, whch tends to result n a better generalzaton of the data. Generalzaton helps wth the common classfcaton problem of over-fttng. The ponts that le on the planes that separate the data are the support vectors. Fndng the support vectors requres solvng the followng optmzaton problem (de-tals of ths method can be found n [2, 13]): mn 1 2 w, b, ξ T subject to: y ( w φ( x ) + b) 1 w l w + C ξ = 1 ξ, ξ > 0 The geometrc representaton of the SVM s easly vsualzed when the data falls nto the lnear separable case. It becomes more complex as the data falls nto lnear nonseparable case and non-lnear separable cases. For more nformaton on the dfferent cases, please refer to [11] whch devotes a chapter on SVMs. Real world data tends to fall nto the non-lnear separable case. To solve the non-lnear separable problem, the SVM T (2) reles on pre-processng the data to represent patterns n a hgher dmenson than the orgnal data set. The functons that provde the mappng to hgher dmensons are known as ph functons or kernels. Common kernels nclude Radal Bass Functon (RBF), lnear, polynomal, and sgmod. The RBF kernel s used n ths study and addtonal nformaton on ths kernel follows n secton 4. The other classfer we consder s the Nearest Neghbor classfer, whch computes dstances from a test subject d to each member d of the database, and classfes the test subject as the subject that has the closest dstance. The dstances can be computed usng varous methods such cty-block dstance or Eucldean dstance. A reject threshold can be ntroduced nto the Nearest Neghbor classfcaton. If the dstance between test subject d and ts nearest neghbor d s wthn the threshold, the classfcaton s that of the closest member. However, f the dstance s greater than the threshold, the subject s rejected and classfed as a non-member. In ths study, we used a reject threshold. 4 EXPERIMENTS Our hypothess s that bometrc dentfcaton on closed systems does not generalze well to larger, open systems contanng non-members. In order to nvestgate ths hypothess, experments were conducted on subset database M ' M from both the wrter and rs databases descrbed n secton Experment Setup For each of the databases, tranng sets were created. Tranng sets for the wrter data conssted of m ' = 50, 100, 200 and 400 members. Tranng sets for the rs data conssted of m ' = 5, 15, 25 and 35 members. These sets ncluded all nstances per member,.e., 3 per member for wrter and 10 per member for rs. For the frst part of the experment, an SVM was traned on the members. Parameter tunng, or SVM optmzaton, was performed pror to tranng. The frst parameter tuned s the penalty parameter C from equaton (2), and dependng on the kernel used, there are addtonal parameters to tune. For ths experment we used an RBF kernel of the form: K( x, xj) = γ x xj, γ > 0 e The γ parameter n equaton (3) s the only kernel parameter requrng tunng. A grd-search method as defned n [8] was used to optmze these two parameters. Tunng the parameters gves 100% accuracy on each of the tranng sets. Therefore, we have 0% FAR and FRR, or equvalently, 100% securty and convenence. The next 2 (3) D5.3

4 step s to test non-members to determne the true securty of the traned SVM. For each tranng set we created a combned evaluaton sets consstng of the traned members plus an ncreasng number of non-members. The evaluaton sets for the 50-wrter traned SVM conssted of 50, 100, 200, 400, 700 and 841 subjects, where the frst 50 subjects are the members and the remanng subjects are non-members. Smlarly, the evaluaton sets for the 25-rs traned SVM conssted of 25, 35, 45 and 52 subjects, where the frst 25 subjects are the members and remanng subjects are non-members. 4.2 Results and Analyss In the postve dentfcaton model we consder two errors: false accepts and false rejects. Snce the SVM was able to tran the members to 100% accuracy, we elmnate the false accepts and false rejects for members. The remanng tests are non-members and therefore can only produce false accepts. The false accepts correlate to the securty measurement of the system, a measure of extreme mportance to the system. In fgure 3, the securty results are shown for the wrter data. In the second part of the experment, the Nearest Neghbor classfer was used. For ths classfer, threshold tunng was requred. The threshold has to be large enough to allow dentfcaton for the known members, but small enough not to allow non-members to be classfed as members. As the threshold ncreases, the FAR ncreases and FRR decreases. The Recever Operatng Characterstc (ROC) curve for the Nearest Neghbor classfer s presented n fgure 2. The ROC curve s a plot of FAR aganst FRR for varous thresholds. When desgnng an dentfcaton system, there s a trade off between the convenence (FRR) and securty (FAR) of the system. For ths experment, we have chosen an operatng threshold that s close to equal error rate, but leanng towards a hgher securty system. Fgure 3. Securty results for wrter data usng SVM Fgure 4. Securty results for rs data usng SVM. Fgure 2. ROC curve for Nearest Neghbor wth 100 members. As hypotheszed, for each curve, as the number of nonmembers ncreases, the securty monotoncally decreases (or equvalently, the FAR monotoncally ncreases). It mght also be noted that the fnal rates to whch the securty curves decrease appear to converge that s, to approach asymptotes. To ensure that ths s not an artfact of the partcular handwrtng data used, we obtaned smlar experment results on the rs data as presented n fgure 4. The rs data n fgure 4 follows the same pattern as the wrter data n fgure 3, although convergence s not as evdent for these data. D5.4

5 Next, we present the results for the Nearest Neghbor classfer. As can be seen n fgure 5, the same pattern emerges, although n ths experment we dd not obtan 0% FAR for the members. When usng the Nearest Neghbor approach, a one-versus-all method was used to obtan the accuracy for the closed envronment. Last, we go one step further brngng n an addtonal classfer, the Artfcal Neural Network or ANN. Fgure 7 llustrates the securty performance for 15 members of the rs database. We notce the Nearest Neghbor agan does not perform well on the closed envronment, but does out perform the SVM performance as non-members enter the system. However, we ntroduced a 1-versus-all approach ANN. Ths approach, lke SVM, provdes excellent performance on the closed envronment, and actually surpasses both the SVM and NN on the open envronment. Ths s an nterestng fnd, however, addtonal work stll needs to performed to test aganst varous szed rs and wrter data. Fgure 5. Securty results for wrter data usng Nearest Neghbor. We now present a comparson of the results from the two classfers used n ths experment. Fgure 6 llustrates the securty performance for 100 members of the wrter database. Notce, although Nearest Neghbor does not perform as well on the closed envronment, t eventually meets and surpasses the performance of the SVM as non-members enter the system. Fgure 7. A comparson of the performance of the Nearest Neghbor, SVM and ANN classfers on the rs data of 15 members. 4.3 Securty Convergence Based on the securty results of fgures 3, 4 and 5, we recognze that the curves appear to be of exponental form and that we mght be able to extrapolate the securty of a system for large populatons contanng non-members. After some fttng trals, we fnd the curve most smlar to be y 1 (( x b) / c) 2 = + ae where the constant b s the number of members; the constants a, c, and d vary based on b; and the securty converges to d. Fgure 8 dsplays the curve fttng for the 50 and 200 traned wrter data samples. d (4) Fgure 6. A comparson of the performance of the Nearest Neghbor and SVM classfers on the wrter data of 100 members. D5.5

6 In summary, we demonstrated that the generalzaton capablty of closed bometrc systems n open envronments s poor, and that the sgnfcantly larger error rates should be taken nto account when desgnng bometrc systems for postve dentfcaton. For ncreased securty, for example, mult-modal bometrcs mght be consdered [14]. 5.1 Future work When desgnng an dentfcaton system, there s a trade off between the convenence and securty of the system. Most systems would choose securty over convenence. However, n our mplementaton of SVM for the wrter data, we mply choosng convenence over securty (guarantee 0 false rejects). In our study on wrter data, there just are not enough samples per member to put nto the testng set. It would therefore be benefcal to run further experments aganst larger bometrc databases. Fgure 8. Curve fttng for wrter data of 50 and 200 members. 5 CONCLUSIONS In ths paper, we found that system securty (1-FAR) decreases rapdly for closed systems when they are tested n open-system mode as the number of non members tested ncreases. Thus, the hgh accuracy rates often obtaned for closed bometrc dentfcaton problems do not appear to generalze well to the open system problem. Ths s mportant because we beleve that no system can be guaranteed to reman closed. Ths hypothess was valdated by experments on both wrter and rs bometrc databases. An estmate of the expected error was also projected based on the asymptote of an exponental curve ftted to the data. Furthermore, because the FAR n these studes was obtaned from normal bometrcs of other subjects, the socalled "zero-effort" rate, the true FAR of these systems would be greater, resultng n even poorer generalzaton. We also found that, although systems can be traned for greater closed-system securty usng SVM rather than NN classfers, the NN systems are better for generalzng to open systems through ther capablty of rejectng non members. Thus, t appears that the reject thresholds of NN classfers do a better job of rejectng non members than the reject regons of SVM classfers. Also, the ANN classfer may provde even better results when generalzng to open systems, but more work stll needs to be performed wth that classfer. Note, most complex bometrcs systems use more complex classfers. Gven that performance on closed systems s not suffcent for applcatons where securty s essental, we feel even the most complex classfers should be tested n an open envronment. For a more n depth analyss of varous classfers on open envronments, refer to [6] We thnk that t may be advantageous to develop open dentfcaton systems by usng a verfcaton model. Also, t would be benefcal to explore the features for the gven bometrcs snce securty results could mprove wth features that better dentfy a member. Addtonal forms of bometrcs, such as fngerprnt, face, voce, hand geometry or a combnaton of bometrcs mght also be tested to further test the hypothess. REFERENCES 1. Bolle, R.M., Connell, J.H., Pankant, S., Ratha, N.K., Senor, A.W.: Gude to Bometrcs. Sprnger (2004) 2. Burges, C.: A Tutoral on Support Vector Machnes for Pattern Recognton. Data Mnng and Knowledge Dscovery. 2: (1998) 3. Cha, S.-H., Srhar, S.N.: Wrter Identfcaton: Statstcal Analyss and Dchotomzer. Proceedngs Internatonal Workshop on Structural and Syntactcal Pattern Recognton (SSPR 2000), Alcante, Span. pp (2000) 4. Cha, S.-H., Srhar, S.N.: Handwrtten Document Image Database Constructon and Retreval System. Proceedngs of SPIE Document Recognton and Retreval VIII, Vol San Jose (2001) 5. Duda, R.O., Hart, P.E., Stork, D.G.: Pattern Classfcaton. John Wley & Sons (2001) 6. Gbbons, M.: On Evaluatng Open Bometrc Identfcaton Systems. Master s Thess, CSIS Department, Pace Unversty (2005) 7. Grother, P., Phllps, P.J.: Models of Large Populaton Recognton Performance IEEE Computer Socety Conference on Computer Vson and Pattern Recognton (CVPR'04) Vol. 2. Washngton, D.C. pp (2004) 8. Hsu, C.-W., Chang, C.-C., Len, C.-J.: A Practcal Gude to Support Vector Classfcaton. 9. Kee, G., Byun, Y., Lee, K., Lee, Y.: Improved Technques for an Irs Recognton System wth Hgh Performance. Lecture Notes Artfcal Intellgence (2001) 10. Krchen, E., Mellakh, M.A., Garca-Salcett, S., Dorzz, B.: Irs Identfcaton Usng Wavelet Packets. Pattern Recognton, 17th Internatonal Conference on (ICPR'04) Vol. 4. pp (2004) D5.6

7 11. Kung, S.Y., Mak, M.W., Ln, S.H.: Bometrc Authentcaton: A Machne Learnng Ap-proach. Pearson Educaton Company (2004) 12. Ma, Y., Pollck, F., Hewtt, W.T.: Usng B-Splne Curves for Hand Recognton. Pattern Recognton, 17th Internatonal Conference on (ICPR'04) Vol. 3. Cambrdge UK. pp (2004) 13. Osuna, E., Freund, R., Gros, F.: Support Vector Machnes: Tranng and Applcatons. MIT Artfcal Intellgence Laboratory and Center for Bologcal and Computatonal Learn-ng Department of Bran and Cogntve Scences. A.I. Memo No 1602, C.B.C.L. Paper No 144 (1997) and Machne Intellgence. Vol. 25, No. 9. pp (2003) 14. Prabhakar, S., Pankant, S., Jan, A.K.: Bometrc Recognton: Securty & Prvacy Con-cerns. IEEE Securty and Prvacy Magazne. Vol. 1, No. 2. pp (2003) 15. Schlapbach, A., Bunke, H.: Off-lne Handwrtng Identfcaton Usng HMM Based Recog-nzers. Pattern Recognton, 17th Internatonal Conference on (ICPR'04) Vol. 2. Cambrdge UK. pp (2004) 16. Woodford, B.J., Deng, D., Benwell, G.L.: A Wavelet-based Neuro-fuzzy System for Data Mnng Small Image Sets. 17. Zhang, D., Kong, W.-K., You, J., Wong, M.: Onlne Palmprnt Identfcaton. IEEE Trans-actons on Pattern Analyss D5.7

Support Vector Machines

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

More information

Machine Learning 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

Lecture 5: Multilayer Perceptrons

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

More information

Classification / Regression Support Vector Machines

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

More information

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

Announcements. Supervised Learning

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

More information

Face Recognition University at Buffalo CSE666 Lecture Slides Resources:

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

More information

Learning the Kernel Parameters in Kernel Minimum Distance Classifier

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

More information

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

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

More information

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

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

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

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

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

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

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

More information

Shape Representation Robust to the Sketching Order Using Distance Map and Direction Histogram

Shape Representation Robust to the Sketching Order Using Distance Map and Direction Histogram Shape Representaton Robust to the Sketchng Order Usng Dstance Map and Drecton Hstogram Department of Computer Scence Yonse Unversty Kwon Yun CONTENTS Revew Topc Proposed Method System Overvew Sketch Normalzaton

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

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

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

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

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

More information

Face Recognition Method Based on Within-class Clustering SVM

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

More information

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

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

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

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

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

PRÉSENTATIONS DE PROJETS

PRÉSENTATIONS DE PROJETS PRÉSENTATIONS DE PROJETS Rex Onlne (V. Atanasu) What s Rex? Rex s an onlne browser for collectons of wrtten documents [1]. Asde ths core functon t has however many other applcatons that make t nterestng

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

Using Neural Networks and Support Vector Machines in Data Mining

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

More information

Fingerprint matching based on weighting method and SVM

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

More information

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

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

Discriminative classifiers for object classification. Last time

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

More information

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

Signature and Lexicon Pruning Techniques

Signature and Lexicon Pruning Techniques Sgnature and Lexcon Prunng Technques Srnvas Palla, Hansheng Le, Venu Govndaraju Centre for Unfed Bometrcs and Sensors Unversty at Buffalo {spalla2, hle, govnd}@cedar.buffalo.edu Abstract Handwrtten word

More information

X- Chart Using ANOM Approach

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

More information

A PATTERN RECOGNITION APPROACH TO IMAGE SEGMENTATION

A 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 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

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

Palmprint Feature Extraction Using 2-D Gabor Filters

Palmprint Feature Extraction Using 2-D Gabor Filters Palmprnt Feature Extracton Usng 2-D Gabor Flters Wa Kn Kong Davd Zhang and Wenxn L Bometrcs Research Centre Department of Computng The Hong Kong Polytechnc Unversty Kowloon Hong Kong Correspondng author:

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

TN348: Openlab Module - Colocalization

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

More information

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

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

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

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

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

More information

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

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

An Image Fusion Approach Based on Segmentation Region

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

More information

Sixth Indian Conference on Computer Vision, Graphics & Image Processing

Sixth Indian Conference on Computer Vision, Graphics & Image Processing Sxth Indan Conference on Computer Vson, Graphcs & Image Processng Incorporatng Cohort Informaton for Relable Palmprnt Authentcaton Ajay Kumar Bometrcs Research Laboratory, Department of Electrcal Engneerng

More information

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

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

More information

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

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

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

More information

SUMMARY... I TABLE OF CONTENTS...II INTRODUCTION...

SUMMARY... I TABLE OF CONTENTS...II INTRODUCTION... Summary A follow-the-leader robot system s mplemented usng Dscrete-Event Supervsory Control methods. The system conssts of three robots, a leader and two followers. The dea s to get the two followers to

More information

Discriminative Dictionary Learning with Pairwise Constraints

Discriminative Dictionary Learning with Pairwise Constraints Dscrmnatve Dctonary Learnng wth Parwse Constrants Humn Guo Zhuoln Jang LARRY S. DAVIS UNIVERSITY OF MARYLAND Nov. 6 th, Outlne Introducton/motvaton Dctonary Learnng Dscrmnatve Dctonary Learnng wth Parwse

More information

Analysis of Continuous Beams in General

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

More information

Feature Extractions for Iris Recognition

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

More information

A Fingerprint Matching Model using Unsupervised Learning Approach

A Fingerprint Matching Model using Unsupervised Learning Approach A Fngerprnt Matchng Model usng Unsupervsed Learnng Approach Nasser S. Abouzakhar and Muhammed Bello Abdulazeez School of Computer Scence, The Unversty of Hertfordshre, College Lane, Hatfeld AL 10 9AB,

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

Parallelism for Nested Loops with Non-uniform and Flow Dependences

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

More information

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

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

More information

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

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

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

More information

An Optimal Algorithm for Prufer Codes *

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

More information

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

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

More information

MATHEMATICS FORM ONE SCHEME OF WORK 2004

MATHEMATICS FORM ONE SCHEME OF WORK 2004 MATHEMATICS FORM ONE SCHEME OF WORK 2004 WEEK TOPICS/SUBTOPICS LEARNING OBJECTIVES LEARNING OUTCOMES VALUES CREATIVE & CRITICAL THINKING 1 WHOLE NUMBER Students wll be able to: GENERICS 1 1.1 Concept of

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

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng, Natonal Unversty of Sngapore {shva, phanquyt, tancl }@comp.nus.edu.sg

More information

SHAPE RECOGNITION METHOD BASED ON THE k-nearest NEIGHBOR RULE

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

More information

Human Face Recognition Using Generalized. Kernel Fisher Discriminant

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

More information

A Gradient Difference based Technique for Video Text Detection

A Gradient Difference based Technique for Video Text Detection 2009 10th Internatonal Conference on Document Analyss and Recognton A Gradent Dfference based Technque for Vdeo Text Detecton Palaahnakote Shvakumara, Trung Quy Phan and Chew Lm Tan School of Computng,

More information

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline

Image Representation & Visualization Basic Imaging Algorithms Shape Representation and Analysis. outline mage Vsualzaton mage Vsualzaton mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and Analyss outlne mage Representaton & Vsualzaton Basc magng Algorthms Shape Representaton and

More information

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields

A mathematical programming approach to the analysis, design and scheduling of offshore oilfields 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 A mathematcal programmng approach to the analyss, desgn and

More information

Training of Kernel Fuzzy Classifiers by Dynamic Cluster Generation

Training 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 information

NUMERICAL SOLVING OPTIMAL CONTROL PROBLEMS BY THE METHOD OF VARIATIONS

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

More information

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

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

On Supporting Identification in a Hand-Based Biometric Framework

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

More information

One-handed Keystroke Biometric Identification Competition

One-handed Keystroke Biometric Identification Competition One-handed Keystroke Bometrc Identfcaton Competton John V. Monaco 1, Gonzalo Perez 1, Charles C. Tappert 1, Patrck Bours 2, Soumk Mondal 2, Sudala Rajkumar 3, Aytham Morales 4, Julan Ferrez 4 and Javer

More information

Related-Mode Attacks on CTR Encryption Mode

Related-Mode Attacks on CTR Encryption Mode Internatonal Journal of Network Securty, Vol.4, No.3, PP.282 287, May 2007 282 Related-Mode Attacks on CTR Encrypton Mode Dayn Wang, Dongda Ln, and Wenlng Wu (Correspondng author: Dayn Wang) Key Laboratory

More information

Local Quaternary Patterns and Feature Local Quaternary Patterns

Local Quaternary Patterns and Feature Local Quaternary Patterns Local Quaternary Patterns and Feature Local Quaternary Patterns Jayu Gu and Chengjun Lu The Department of Computer Scence, New Jersey Insttute of Technology, Newark, NJ 0102, USA Abstract - Ths paper presents

More information

y and the total sum of

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

More information

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

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

On Some Entertaining Applications of the Concept of Set in Computer Science Course

On Some Entertaining Applications of the Concept of Set in Computer Science Course On Some Entertanng Applcatons of the Concept of Set n Computer Scence Course Krasmr Yordzhev *, Hrstna Kostadnova ** * Assocate Professor Krasmr Yordzhev, Ph.D., Faculty of Mathematcs and Natural Scences,

More information

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

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

More information

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

PERFORMANCE EVALUATION FOR SCENE MATCHING ALGORITHMS BY SVM

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

More information

An AAM-based Face Shape Classification Method Used for Facial Expression Recognition

An AAM-based Face Shape Classification Method Used for Facial Expression Recognition Internatonal Journal of Research n Engneerng and Technology (IJRET) Vol. 2, No. 4, 23 ISSN 2277 4378 An AAM-based Face Shape Classfcaton Method Used for Facal Expresson Recognton Lunng. L, Jaehyun So,

More information

Japanese Dependency Analysis Based on Improved SVM and KNN

Japanese Dependency Analysis Based on Improved SVM and KNN Proceedngs of the 7th WSEAS Internatonal Conference on Smulaton, Modellng and Optmzaton, Bejng, Chna, September 15-17, 2007 140 Japanese Dependency Analyss Based on Improved SVM and KNN ZHOU HUIWEI and

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

Web Spam Detection Using Multiple Kernels in Twin Support Vector Machine

Web Spam Detection Using Multiple Kernels in Twin Support Vector Machine Web Spam Detecton Usng Multple Kernels n Twn Support Vector Machne ABSTRACT Seyed Hamd Reza Mohammad, Mohammad Al Zare Chahook Yazd Unversty, Yazd, Iran mohammad_6468@stu.yazd.ac.r chahook@yazd.ac.r Search

More information

Support Vector Machines. CS534 - Machine Learning

Support Vector Machines. CS534 - Machine Learning Support Vector Machnes CS534 - Machne Learnng Perceptron Revsted: Lnear Separators Bnar classfcaton can be veed as the task of separatng classes n feature space: b > 0 b 0 b < 0 f() sgn( b) Lnear Separators

More information

Lecture #15 Lecture Notes

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

More information

3D vector computer graphics

3D vector computer graphics 3D vector computer graphcs Paolo Varagnolo: freelance engneer Padova Aprl 2016 Prvate Practce ----------------------------------- 1. Introducton Vector 3D model representaton n computer graphcs requres

More information

On Modeling Variations For Face Authentication

On Modeling Variations For Face Authentication On Modelng Varatons For Face Authentcaton Xaomng Lu Tsuhan Chen B.V.K. Vjaya Kumar Department of Electrcal and Computer Engneerng, Carnege Mellon Unversty Abstract In ths paper, we present a scheme for

More information

Wavelets and Support Vector Machines for Texture Classification

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

More information

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

Comparison Study of Textural Descriptors for Training Neural Network Classifiers

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

More information

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