Biometric Signature Processing & Recognition Using Radial Basis Function Network
|
|
|
- Nathan Caldwell
- 10 years ago
- Views:
Transcription
1 Bometrc Sgnature Processng & Recognton Usng Radal Bass Functon Network Ankt Chadha, Neha Satam, and Vbha Wal Abstract- Automatc recognton of sgnature s a challengng problem whch has receved much attenton durng recent years due to ts many applcatons n dfferent felds. Sgnature has been used for long tme for verfcaton and authentcaton purpose. Earler methods were manual but nowadays they are gettng dgtzed. Ths paper provdes an effcent method to sgnature recognton usng Radal Bass Functon Network. The network s traned wth sample mages n database. Feature extracton s performed before usng them for tranng. For testng purpose, an mage s made to undergo rotaton-translaton-scalng correcton and then gven to network. The network successfully dentfes the orgnal mage and gves correct output for stored database mages also. The method provdes recognton rate of approxmately 80% for 200 samples. Keywords- Database, Feature extracton, Radal Bass Functon Network, Sgnature recognton. I I. INTRODUCTION n everyday lfe there are many places where we are needed to gve dentfcaton to gan access. Its examples nclude nternet account, credt cards, ATM machnes, etc. We have to frst provde our dentfcaton and then we can access them. These dentfcatons are generally a logn ID and an alphanumerc (maybe wth specal characters) password or a Personal Identfcaton Number (PIN). Man drawback of ths system s that, they provde verfcaton and not actually dentfcaton. These two terms, though used nterchangeably, are dfferent. When a system verfes a password, t ust checks whether the password s rght or wrong. It doesn t dentfy whch person s ganng access. Hence, anyone havng knowledge of another person s password can easly get hs/her data and can tamper t. Wth the advent of new technology, these systems are beng replaced by much more advanced technques to dentfy a person. These technques are called bometrcs, whch nvolve checkng a person s bologcal trats such as face, retna, fngerprnt, rs, voce, sgnature etc. Formally, bometrcs refers to the dentfcaton of humans by ther characterstcs or trats. The system already has the samples Manuscrpt receved September 30, A. R. Chadha s wth the Vdyalankar Insttute of Technology, Mumba, , INDIA (phone: ; emal:[email protected]). N. S. Satam s wth the Vdyalankar Insttute of Technology, Mumba, , INDIA (phone: ; emal:[email protected]). V. S. Wal s an Assstant Professor n the Electroncs and Telecommuncatons Engneerng Department, Vdyalankar Insttute of Technology, Mumba, , INDIA (e-mal: [email protected]). of characterstcs of all users. When a person provdes hs/her dentty, say fngerprnt, system checks t wth the records n database. When any of records match wth the one beng provded for authentcaton, the user s dentfed by system & gven access. Some of the applcatons of bometrc recognton nclude drvng lcenses, mmgraton, natonal ID, passport, voter regstraton, securty applcatons, medcal records, personal devce logon, desktop logon, human-robot nteracton, human-computer nteracton, smart cards etc. There are certan bologcal measurements whch qualfy to be a bometrc. As sad above, face, retna, rs, sgnature are used as bometrc characterstcs because they are dfferent for dfferent people. Followng are some requrements [1] whch are needed to be satsfed n order to use them as bometrc characterstcs: 1. Unversalty: everyone should have the characterstc. 2. Unqueness: two people should have some dfference n characterstc. 3. Durablty: the characterstc should reman same over the tme. 4. Collectvely: the characterstc should be avalable n quantty. Though above are the propertes of characterstcs, practcally, we have to consder some other ssues also [1]: 1. Performance: the characterstcs should gve excellent speed and accuracy wth mnmum explotaton of resources. 2. Acceptablty: people should accept the use of characterstc n ther daly lfe. 3. Crcumventon: represents how easly the system can be deceved. A plan bometrc system s equpped wth sensory unt, a unt for feature extracton, a unt to perform the matchng and decson-makng unt. The sensory unt collects the bometrc data of user. For the sgnature recognton purpose, a dgtal tablet along wth pen-lke stylus s provded whch works as sensor. The promnent features of data are extracted usng feature extracton unt. These features sgnfy the orgnal data but n usng lesser attrbutes. They can be poston and orentaton of certan ponts, curved lnes n sgnature, underlned alphabets etc. Then ths feature set s matched aganst every set stored n the database and a matchng score s generated. Ths score
2 depends on how many ponts are matched between data n database and newly produced data. Once ths score s obtaned then t s responsblty of decson-makng unt whether to accept the ndvdual as dentfed user or not. The age-old method for sgnature verfcaton s manual one, n whch a person manually checks the sgnature. If both the sgnatures are suffcently smlar, then the access s granted. An automated sgnature verfcaton process wll help mprove the scenaro. Hgh-end bometrc methods nclude rs, retna, face and fngerprnt based dentfcaton and verfcaton. Even though human features such as rs, retna and fngerprnts are fxed for a person and the varatons n the respectve bometrc attrbute are low, specal and relatvely expensve hardware s needed for data acquston n such systems. For the varous purposes such as bankng, fnancal transactons, document authentcaton, sgnature s used snce long tme. Also, the verfcaton s relatvely less expensve than other bometrc systems [2]. Though the verfcaton s effcent, t s dffcult. There are some varatons n a person's sgnature over the tme, whch needs to be accounted. Examples of the varous varatons observed n the sgnature of an ndvdual have been llustrated n Fg. 1. Every tme sgnature s made, t dffers from prevous one n some angle or scale or curvng of letters. Fg 1: Varatons n the sgnature of an ndvdual Ths nvolves elmnatng or reducng the rotaton, scalng and translaton factors between the orgnal and the newly produced sgnatures. For the verfcaton & dentfcaton purpose, neural network can be employed. Fg.2 shows block dagram of proposed system, where all the effects of new sgnature are reduced and t s matched aganst those n database usng neural network. New Image Rotaton, Scalng & Translaton removal Feature extracton usng DCT Radal Bass Functon Network Verfcaton Database consstng of 800 samples Feature extracton usng DCT Fg.2: Block dagram of proposed system Frst, new mage, also known as user mage s made to undergo Rotaton-Scalng-Translaton removal, so that t resembles orgnal sgnature stored n database. Database s created by collectng 800 sgnature samples. These sgnatures are stored n database after feature extracton process for whch DCT s employed. It s also appled on corrected user mage. Now ths mage s provded to Radal Bass Functon Network (RBFN), whch s earler traned usng database. If RBFN dentfes user mage as the one n database then access for that partcular ndvdual s granted. The advantage of ths system s that t uses neural network effectvely,.e., t requres only few samples to tran the network and then t performs for sgnatures newly added n the database. The algorthm used n ths paper uses correlaton to detect the rotaton and mplements smple croppng method to elmnate the effect. RBFN s used for verfcaton purpose. In paper [3], back-propagaton algorthm s used for sgnature recognton. In ths algorthm, neural network s created wth R,G and B pxel values as nput and gray value as output. To calculate the gradent, unsgned char to unsgned nt and float to unsgned char conversons are needed to be performed. Ths ncreases computaton. Also, the network s traned multple tmes so that proper weght matrx s obtaned. Use of RBFN reduces ths need of multple tranng and acheves the goal n few epochs. Ths paper s organzed as follows: Secton II descrbes dea of proposed soluton. Secton III brefs mplementaton steps and secton IV shows expermental results. Secton V provdes concluson.
3 II. IDEA OF PROPOSED SOLUTION The prmary concern s decdng at what angle the new sgnature devates from the sgnature n database. For ths, concept of correlaton can be mplemented. Correlaton s a statstcal measure, whch refers to a process for establshng whether or not relatonshps exst between two varables [4]. It denotes the value by whch two varable vary. If the value s maxmum then there s maxmum smlarty n the two varables. Consder orgnal sgnature as varable P ( m, n) and new sgnature to be tested as varable Q ( m, n). If P 0 and 0 Q are mean values of P and Q then cross-correlaton r between them s descrbed by followng equaton: r P P )( Q Q ) m n ( mn 0 mn 0 Normalzed cross- correlaton can be used to smplfy examnaton and comparsons of coeffcent values correspondng to the respectve angular values. Mn-max normalzaton s the procedure used to obtan normalzed cross - correlaton [5]. Mn - max normalzaton mantans the relatonshps among the orgnal data values. The normalzaton operaton transforms the data nto a new range, generally [0, 1]. Gven a data set x, such that 1,2, n, the normalzed value followng equaton: (1) ' x mn( x ) x (2) max( x ) mn( x ) ' x s gven by the The second goal s to obtan the translaton assocated wth the mages. Ths can be accomplshed by a smple croppng technque. Intally, the number of rows and columns borderng the sgnature pxels wthn the mage are calculated. The mage devod of these rows and columns s extracted. The result s an mage consstng of only the sgnature pxels. Ths elmnates addtonal background surroundng the mage. Thrd goal s scalng between two mages. For calculaton of the scalng factor, the cropped mages obtaned durng translaton are used. The sze of the orgnal mage dvded by the sze of the user mage gves the scalng rato. tp such that only pen- down samples (.e., when the pen touches the paper) are recorded. Fg. 3: Wacom Bamboo Dgtal Pen Tablet For testng the system, a database was created. It conssted of a set of sgnature samples of 100 people. For each person, there are 9 test samples and 1 tranng sample. The samples are color normalzed. Color normalzaton converts the colored mage nto correspondng grayscale mage. B. Rotaton Elmnaton There s always some tltng of sgnatures when there s no reference lne provded for t. Ths angle of tlt vares from - 60 o to +60 o. Before verfcaton process, the rotaton s neutralzed by algnng the new mage wth orgnal mage. The new mage s then rotated by 5 wthn the range of 60 to +60 n successve teratons. Cross-correlaton values between the reference mage and the new mage are recorded on completon of each teraton of the rotaton process. The maxmum cross-correlaton value refers to the correct angle of rotaton wthn a 5 range, further, after the approxmate angle value s obtaned, +3 or 3 of ths angle can be nspected for maxmum correlaton value whch corresponds to angle of rotaton accurate to up to 1.The user mage s rotated by the negatve of the angle and we get the mage whch s free from rotaton. Fg. 4 gves the flowchart of ths algorthm. A. Sample Acquston & Pre-processng of Images The system mplemented here uses a dgtal pen tablet, namely, WACOM Bamboo shown n fg.3, as the datacapturng devce. The pen has a touch senstve swtch n ts
4 Orgnal Image Pre-processed orgnal mage Startng from θ = -60 o START Color normalzaton Removal of background New Image User mage Calculaton of normalzed correlaton value Is θ 60 o? Yes No translaton n X and/or Y drecton, havng a maxmum value equal to the wdth or heght of the sgnature canvas respectvely. The soluton s gven by croppng method n whch only the sgnature pxels are extracted. So wth respect to new mage formed, there s no such translaton n X and/or Y drecton. Ths elmnates the effect of translaton. The number of columns from left and number of rows from the bottom, whch contan no black pxels correspondng to the actual sgnature,.e., whch consst solely of mage- background, are counted. These values gve X translaton and Y translaton respectvely. D. Scalng Elmnaton The sze of the sgnature vares accordng to the sze of space. For smaller spaces, the sgnature may be compressed, for larger space, the sgn may be enlarged. Thus, before extracton of features, t s essental that any scalng, f present n the test sample, be removed. Upon croppng both mages, the rato of heght gves Y scalng and rato of wdth gves X scalng. However, to resze the user mage and make t the same sze as the mage, ether of the scalng ratos can be used. For a rotaton range of 60 to +60, heght was observed to vary sgnfcantly as compared to the length. Hence, Y scalng was chosen as the scalng rato. Scalng rato s calculated by the followng equaton: Scalng rato Sze of reference mage Sze of test mage Increment θ by 5 o No Is Correlaton maxmum? Yes Correct the user mage by θ o The user mage s reszed as per the obtaned scalng rato and then sent to the feature extracton segment. E. Combned Rotaton Scalng Translaton It s not dffcult to control the specmens to get pure rotaton, translaton and scalng, n any case, for orgnal sgnature, all the aforementoned components are adusted synchronously. Thus, rotaton, translaton and scalng modfcatons are connected n the same manner. Fg.4: Flowchart of algorthm C. Translaton Elmnaton STOP Smlar to tltng of sgnature, there s one more dffculty whch s translaton. It s the effect whch ntroduces Rotaton revson goes before translaton revson as the expected cause at base left corner lkewse gets rotated and translaton mpacts can't be dspensed wth unless the orgn s rotated back to bottom left as correctly as could be allowed. In ths way, rotaton adustment needs to be performed frst as the scalng proporton computed by the pure scalng strategy s not relable wth scalng degree of the rotated mage.
5 Hence, the effcency of scalng revson depends, to a substantal degree, on the percentage error obtaned durng rotaton correcton. archtecture of RBFN has proved to drectly mprove tranng and performance of the network [8]-[9]. Fg. 5 depcts archtecture of RBFN. After correctng the angle of rotaton, the user mage preprocessed duplcate s edted to take out translaton and the mage so acqured s an nstance of pure scalng whch has been dscussed prevously. Along these lnes, rotaton scalng translaton cancellaton s performed. F. Feature Extracton usng DCT DCT s one of the most wdely used transform n the used for feature extracton. It nvolves takng the transformaton of the mage as a whole and separatng the relevant coeffcents. DCT performs energy compacton [6]. The DCT of an mage bascally conssts of three frequency components namely low, mddle, hgh each contanng some detal and nformaton n an mage. The low frequency generally contans the average ntensty of an mage whch s the most ntended n FR systems [7]. Mathematcally, the 2D-DCT of an mage s gven by: F( u, v) ( u) ( v) ( u) ( v) N 1 M 1 x0 y0 1 N 2 N u u cos (2x 1) cos (2y 1) f ( x, y) 2N 2M for u, v 0 for u, v 0 (3) where f ( x, y) s the ntensty of the pxel at coordnates ( x, y), u vares from 0 to M-1, and v vares from 0 to N-1, where M N s the sze of mage. G. Radal Bass Functon Network (RBFN) The constructon of RBFN n ts basc from nvolves three layers, vz., nput layer, hdden layer and output layer. All these layers are assgned dfferent roles. The nput layer s formed by sensory unts that connect the network to ts surroundng. The second layer, the hdden layer n entre network, apples nonlnear transformaton from the nput space to hdden space, where dmensonalty of hdden space s larger than that of nput space. Each neuron n hdden layer has a specal type of actvaton functon centered on the center vector of a cluster or subcluster n the feature space so that the functon has non neglgble response for nput vectors close to ts center. Output layer s lnear, producng response of the network to the actvaton sgnals appled to nput layer. Ths partcular Fg. 5: General archtecture of RBFN The learnng process undertaken by an RBFN can be descrbed as follows. The lnear weghts assocated wth the output unt(s) of the network tend to evolve on a dfferent tme scale compared to nonlnear actvaton functons of the hdden unt. Thus as hdden layer's actvaton functons evolve slowly n accordance wth some nonlnear optmzaton strategy; the output layer's weghts adust themselves rapdly through a lnear optmzaton strategy. There are dfferent learnng strateges that can be followed, based on partcular applcaton. It depends on how centers of RBFN are specfed. For ths applcaton of sgnature recognton, approach of supervsed selecton of centers s taken. Several learnng algorthms have been proposed n the lterature for tranng RBF networks [10] [15]. Selecton of a learnng algorthm for a partcular applcaton s crtcally dependent on ts accuracy and speed. In practcal onlne applcatons, sequental learnng algorthms are generally preferred over batch learnng algorthms as they do not requre retranng whenever a new data s receved. In ths approach, centers of RBFN and all other free parameters of the network undergo a supervsed learnng process. It means that RBFN takes ts most generalzed form of error-correcton learnng. The frst step n the development of such a learnng procedure s to defne the nstantaneous value of the cost functon [16]. ξ 1 2 N 2 e 1 (3)
6 Where N s the sze of tranng sample used for learnng and e s the error sgnal defned by e d d * F ( x ) M 1 G x t (4) C The requrement s to fnd the free parameters, t and 1 (the latter beng related to norm-weghtng matrx C ) so as to mnmze ξ. III. A. Image Database IMPLEMENTATION STEPS TABLE 1 COMBINED ROTATION SCALING TRANSLATION Sample Orgnal Parameters Detected Parameters Rotaton Scalng Rotaton Scalng For the expermental purpose, a database of 700 samples was created by collectng 10 samples each from 70 people. The network for dentfcaton was created by usng RBFN fewer neurons model. All the mages were used for tranng of the network. One of the mages was processed by usng rotaton-translaton-scalng algorthm. Fg.6 shows some of the sgnatures used n tranng and testng mage database constructed. Fg.6: Some samples from sgnature database The sgnature recognton system presented n ths paper was developed, traned, and tested usng MATLAB 7. The computer was a Wndows 8 machne wth a 2.5 GHz Intel Core I5 processor and 4 GB of RAM. B. Valdaton of technque The preprocessed grayscale mages of sze 8 8 pxels are reformed n MATLAB to form a 64 1 array wth 64 rows and 1 column for each mage. Ths technque s performed on all test mages to form the nput data for testng the recognton system. Smlarly, the mage database for tranng uses 50 mages and forms a matrx of wth 64 rows and 50 columns. The nput vectors defned for the RBFN are dstrbuted over a 2D-nput space varyng over [0 255], whch represents ntensty levels of the grayscale pxels. As many as 5 test mages are used wth the mage database for performng the experments. Tranng and testng sets were used wthout any overlappng. IV. EXPERIMENTAL RESULTS A. Combned Rotaton Scalng Translaton Table 1 dsplays result for combned rotaton scalng translaton n orgnal rotaton and scalng and detected rotaton and scalng. TABLE 1 COMBINED ROTATION SCALING TRANSLATION The graphcal plot of error n actual and detected rotaton angle s depcted n fgure 7 below.
7 Error percentage Error percentage CT Internatonal Journal of Dgtal Image Processng, ISSN (Prnt) & ISSN X (Onlne) B. Recognton usng MATLAB The network was traned usng varable number of samples. The performance graph of network for 700 samples s shown n fgure Sample Fg. 7: Graphcal plot of error n actual and detected rotaton angle The graphcal plot of error n actual and detected scalng parameter can be seen n fgure 8 below Fg. 12: Performance graph of network for 700 samples The test mage gven to network s shown n fgure 13. The network was successfully able to recognze t though the mage was tlted Sample Fg. 8: Graphcal plot of error n actual and detected scalng parameter Fg. 13: User Image Fg.14: Image recognzed Table 2 dsplays convergence rates relatng to dfferent Mean Square Errors (MSE) and spreads. Orgnal Image Fg.10: Image after Rotaton Correcton Fg.11: Image after Scalng Correcton Fg.9: TABLE 2 CONVERGENCE RATES RELATING TO DIFFERENT MEAN SQUARE ERRORS (MSE) AND SPREADS MSE error Spread Iteratons e e
8 For varous numbers of samples n tranng set, the recognton rate of system also vared. It s tabulated n table. TABLE 3 RECOGNITION RATE FOR VARIOUS NUMBERS OF SAMPLES Number of samples Recognton rate % % % % % V. CONCLUSION Ths paper presents a novel method to sgnature recognton usng RBFN. The system was evaluated n MATLAB usng an mage database of 700 sgnatures, contanng 70 people and each person wth 10 sgnatures. After tranng for approxmately 200 samples the system acheved a recognton rate of 80%. A reduced feature space substantally reduces the computatonal requrements of the method as compared wth standard DCT feature extracton methods. Ths makes our system well suted for low-cost, real-tme hardware mplementaton. Commercal mplementatons of ths technque do not currently exst. However, t s concevable that a practcal RBFN-based sgnature recognton system may be possble n the future. VI. FUTURE WORK The system s mage-based and can be extended to real tme sgnature recognton also. Wth IOS and Androd based systems, touch screen devces lke tablets, phablets and smartphones can be effectvely used for dgtal sgnature dentfcaton and authentcaton. Genetc algorthm can be utlzed to solve ths problem of verfcaton wth reduced computaton and ncreased effcency. VII. REFERENCES [1] Anl Jan, Arun Ross and Sall Prabhakar, Introducton to Bometrc Recognton, IEEE Transactons on Crcuts and Systems for Vdeo Technology, Vol. 14, No. 1, January 2004, pp [2] Anl Jan, F. D. Gress and S. D. Connell, On-lne sgnature verfcaton, Elsever, Pattern Recognton 35, 2002, pp [3] Debnath Bhattacharyya1 and Ta-Hoon Km, Sgnature Recognton usng Artfcal Neural Network, Proc. 9 th WSEAS Internatonal Conference on Computatonal Intellgence, Man-Machne Systems and Cybernetcs (CIMMACS 10), 2010., pp [4] R. J. Rummel, "Understandng Correlaton", Honolulu: Department of Poltcal Scence, Unversty of Hawa, [5] J. Han, M. Kamber, "Data mnng: concepts and technques," Morgan Kaufmann, 2006, pp [6] N. Ahmed, T. Nataraan, K.R. Rao, Dscrete Cosne Transform, IEEE Transactons on Computers, Vol. C-23 (1974) [7] Dgtal Image Processng, Rafael Gonzalez, Rchard Woods, Edton 3, Pearson Publcatons, [8] Song, Y.H., Xuan, Q.Y., and Johns, A.T., Protecton Scheme for E H V Transmsson Systems wth Thyrstor Controlled Seres Compensaton Usng Radal Bass Functon Neural Networks, Electr. Mach. Power Syst., 1997, 25, pp [9] Zohre Fashfar and Javad Haddadna, Desgnng an Fuzzy RBF Neural Network wth Optmal Number of Neuron n Hdden Layer & Effect of Sgnature Shape for Persan Sgnature Recognton by Zernke Moments and PCA, Journal of Computatonal Informaton Systems, September, 2010, 2010 Bnary Informaton Press. [10] L. Yngwe, N. Sundararaan, and P. Saratchandran, A sequental learnng scheme for functon approxmaton usng mnmal radal bass functon (RBF) neural networks, Neural Computat., vol. 9, pp , [11] S. Chen, C. F. N. Cowan, and P. M. Grant, Orthogonal least squares learnng algorthm for radal bass functon networks, IEEE Trans. Neural Netw., vol. 2, no. 2, pp , Mar [12] J. Platt, A resource-allocatng network for functon nterpolaton, Neural Computat., vol. 3, pp , [13] N. B. Karayanns and G. W. M, Growng radal bass neural networks: Mergng supervsed and unsupervsed learnng wth network growth technques, IEEE Trans. Neural Netw., vol. 8, no. 6, pp , Nov [14] M. Salmerón, J. Ortega, C. G. Puntonet, and A. Preto, Improved RAN sequental predcton usng orthogonal technques, Neurocomput., vol. 41, pp , [15] I. Roas, H. Pomares, J. L. Berner, J. Ortega, B. Pno, F. J. Pelayo, and A. Preto, Tme seres analyss usng normalzed PG-RBF network wth regresson weghts, Neurocomput., vol. 42, pp , [16] Smon Haykns, "Neural Networks A Comprehensve Foundaton", Second Edton, Pearson Educaton Inc., 2005, pp
9 Ankt R. Chadha (M 2010) was born n Mumba (M.H.) n Inda on November 07, He s currently pursung hs undergraduate studes n the Electroncs and Telecommuncaton Engneerng dscplne at Vdyalankar Insttute of Technology Mumba. Hs specal felds of nterest nclude Image Processng, Computer Vson (partcularly, Pattern Recognton) and Embedded Systems. He has 6 papers n Internatonal Journals & Conferences to hs credt. Neha S. Satam (M 2010) was born n Mumba (M.H.) n Inda on November 26, She s currently pursung her undergraduate studes n the Electroncs and Telecommuncaton Engneerng dscplne at Vdyalankar Insttute of Technology, Mumba. Her felds of nterest nclude Image Processng, Stegnography and Embedded systems. She has 6 papers n Internatonal Journals & Conferences to her credt. Prof. Vbha S. Wal (M 1997) has done B.E. n Electroncs n 1997 from Shva Unversty and Masters n Electroncs and Telecommuncaton from Mumba Unversty. She had taught subects lke computer graphcs, obect orented programmng, data structure, Image processng and many more. She s currently workng as Assstant professor n Vdyalankar college of Technology, Wadala n Informaton Technology department and presently teachng subects lke Electromagnetc Wave Theory, Electromagnetc Engneerng, Neural network and Fuzzy Logc and Rado Frequency Crcut Desgn. Her area of nterest s Image processng and Fractal Image compresson. She has guded numerous undergraduate proects and papers.
Statistical Approach for Offline Handwritten Signature Verification
Journal of Computer Scence 4 (3): 181-185, 2008 ISSN 1549-3636 2008 Scence Publcatons Statstcal Approach for Offlne Handwrtten Sgnature Verfcaton 2 Debnath Bhattacharyya, 1 Samr Kumar Bandyopadhyay, 2
The Development of Web Log Mining Based on Improve-K-Means Clustering Analysis
The Development of Web Log Mnng Based on Improve-K-Means Clusterng Analyss TngZhong Wang * College of Informaton Technology, Luoyang Normal Unversty, Luoyang, 471022, Chna [email protected] Abstract.
Vision Mouse. Saurabh Sarkar a* University of Cincinnati, Cincinnati, USA ABSTRACT 1. INTRODUCTION
Vson Mouse Saurabh Sarkar a* a Unversty of Cncnnat, Cncnnat, USA ABSTRACT The report dscusses a vson based approach towards trackng of eyes and fngers. The report descrbes the process of locatng the possble
Forecasting the Demand of Emergency Supplies: Based on the CBR Theory and BP Neural Network
700 Proceedngs of the 8th Internatonal Conference on Innovaton & Management Forecastng the Demand of Emergency Supples: Based on the CBR Theory and BP Neural Network Fu Deqang, Lu Yun, L Changbng School
Face Verification Problem. Face Recognition Problem. Application: Access Control. Biometric Authentication. Face Verification (1:1 matching)
Face Recognton Problem Face Verfcaton Problem Face Verfcaton (1:1 matchng) Querymage face query Face Recognton (1:N matchng) database Applcaton: Access Control www.vsage.com www.vsoncs.com Bometrc Authentcaton
A Secure Password-Authenticated Key Agreement Using Smart Cards
A Secure Password-Authentcated Key Agreement Usng Smart Cards Ka Chan 1, Wen-Chung Kuo 2 and Jn-Chou Cheng 3 1 Department of Computer and Informaton Scence, R.O.C. Mltary Academy, Kaohsung 83059, Tawan,
What is Candidate Sampling
What s Canddate Samplng Say we have a multclass or mult label problem where each tranng example ( x, T ) conssts of a context x a small (mult)set of target classes T out of a large unverse L of possble
Module 2 LOSSLESS IMAGE COMPRESSION SYSTEMS. Version 2 ECE IIT, Kharagpur
Module LOSSLESS IMAGE COMPRESSION SYSTEMS Lesson 3 Lossless Compresson: Huffman Codng Instructonal Objectves At the end of ths lesson, the students should be able to:. Defne and measure source entropy..
Forecasting the Direction and Strength of Stock Market Movement
Forecastng the Drecton and Strength of Stock Market Movement Jngwe Chen Mng Chen Nan Ye [email protected] [email protected] [email protected] Abstract - Stock market s one of the most complcated systems
Feature selection for intrusion detection. Slobodan Petrović NISlab, Gjøvik University College
Feature selecton for ntruson detecton Slobodan Petrovć NISlab, Gjøvk Unversty College Contents The feature selecton problem Intruson detecton Traffc features relevant for IDS The CFS measure The mrmr measure
A DATA MINING APPLICATION IN A STUDENT DATABASE
JOURNAL OF AERONAUTICS AND SPACE TECHNOLOGIES JULY 005 VOLUME NUMBER (53-57) A DATA MINING APPLICATION IN A STUDENT DATABASE Şenol Zafer ERDOĞAN Maltepe Ünversty Faculty of Engneerng Büyükbakkalköy-Istanbul
Gender Classification for Real-Time Audience Analysis System
Gender Classfcaton for Real-Tme Audence Analyss System Vladmr Khryashchev, Lev Shmaglt, Andrey Shemyakov, Anton Lebedev Yaroslavl State Unversty Yaroslavl, Russa [email protected], [email protected], [email protected],
Frequency Selective IQ Phase and IQ Amplitude Imbalance Adjustments for OFDM Direct Conversion Transmitters
Frequency Selectve IQ Phase and IQ Ampltude Imbalance Adjustments for OFDM Drect Converson ransmtters Edmund Coersmeer, Ernst Zelnsk Noka, Meesmannstrasse 103, 44807 Bochum, Germany [email protected],
Lecture 2: Single Layer Perceptrons Kevin Swingler
Lecture 2: Sngle Layer Perceptrons Kevn Sngler [email protected] Recap: McCulloch-Ptts Neuron Ths vastly smplfed model of real neurons s also knon as a Threshold Logc Unt: W 2 A Y 3 n W n. A set of synapses
An interactive system for structure-based ASCII art creation
An nteractve system for structure-based ASCII art creaton Katsunor Myake Henry Johan Tomoyuk Nshta The Unversty of Tokyo Nanyang Technologcal Unversty Abstract Non-Photorealstc Renderng (NPR), whose am
An Interest-Oriented Network Evolution Mechanism for Online Communities
An Interest-Orented Network Evoluton Mechansm for Onlne Communtes Cahong Sun and Xaopng Yang School of Informaton, Renmn Unversty of Chna, Bejng 100872, P.R. Chna {chsun,yang}@ruc.edu.cn Abstract. Onlne
A Novel Methodology of Working Capital Management for Large. Public Constructions by Using Fuzzy S-curve Regression
Novel Methodology of Workng Captal Management for Large Publc Constructons by Usng Fuzzy S-curve Regresson Cheng-Wu Chen, Morrs H. L. Wang and Tng-Ya Hseh Department of Cvl Engneerng, Natonal Central Unversty,
An Enhanced Super-Resolution System with Improved Image Registration, Automatic Image Selection, and Image Enhancement
An Enhanced Super-Resoluton System wth Improved Image Regstraton, Automatc Image Selecton, and Image Enhancement Yu-Chuan Kuo ( ), Chen-Yu Chen ( ), and Chou-Shann Fuh ( ) Department of Computer Scence
A hybrid global optimization algorithm based on parallel chaos optimization and outlook algorithm
Avalable onlne www.ocpr.com Journal of Chemcal and Pharmaceutcal Research, 2014, 6(7):1884-1889 Research Artcle ISSN : 0975-7384 CODEN(USA) : JCPRC5 A hybrd global optmzaton algorthm based on parallel
VRT012 User s guide V0.1. Address: Žirmūnų g. 27, Vilnius LT-09105, Phone: (370-5) 2127472, Fax: (370-5) 276 1380, Email: info@teltonika.
VRT012 User s gude V0.1 Thank you for purchasng our product. We hope ths user-frendly devce wll be helpful n realsng your deas and brngng comfort to your lfe. Please take few mnutes to read ths manual
APPLICATION OF PROBE DATA COLLECTED VIA INFRARED BEACONS TO TRAFFIC MANEGEMENT
APPLICATION OF PROBE DATA COLLECTED VIA INFRARED BEACONS TO TRAFFIC MANEGEMENT Toshhko Oda (1), Kochro Iwaoka (2) (1), (2) Infrastructure Systems Busness Unt, Panasonc System Networks Co., Ltd. Saedo-cho
Causal, Explanatory Forecasting. Analysis. Regression Analysis. Simple Linear Regression. Which is Independent? Forecasting
Causal, Explanatory Forecastng Assumes cause-and-effect relatonshp between system nputs and ts output Forecastng wth Regresson Analyss Rchard S. Barr Inputs System Cause + Effect Relatonshp The job of
On-Line Fault Detection in Wind Turbine Transmission System using Adaptive Filter and Robust Statistical Features
On-Lne Fault Detecton n Wnd Turbne Transmsson System usng Adaptve Flter and Robust Statstcal Features Ruoyu L Remote Dagnostcs Center SKF USA Inc. 3443 N. Sam Houston Pkwy., Houston TX 77086 Emal: [email protected]
Traffic-light a stress test for life insurance provisions
MEMORANDUM Date 006-09-7 Authors Bengt von Bahr, Göran Ronge Traffc-lght a stress test for lfe nsurance provsons Fnansnspetonen P.O. Box 6750 SE-113 85 Stocholm [Sveavägen 167] Tel +46 8 787 80 00 Fax
Offline Verification of Hand Written Signature using Adaptive Resonance Theory Net (Type-1)
Internatonal Journal of Sgnal Processng Systems Vol, No June 203 Offlne Verfcaton of Hand Wrtten Sgnature usng Adaptve Resonance Theory Net (Type-) Trtharaj Dash Veer Surendra Sa Unversty of Technology,
A Multi-Camera System on PC-Cluster for Real-time 3-D Tracking
The 23 rd Conference of the Mechancal Engneerng Network of Thaland November 4 7, 2009, Chang Ma A Mult-Camera System on PC-Cluster for Real-tme 3-D Trackng Vboon Sangveraphunsr*, Krtsana Uttamang, and
benefit is 2, paid if the policyholder dies within the year, and probability of death within the year is ).
REVIEW OF RISK MANAGEMENT CONCEPTS LOSS DISTRIBUTIONS AND INSURANCE Loss and nsurance: When someone s subject to the rsk of ncurrng a fnancal loss, the loss s generally modeled usng a random varable or
RESEARCH ON DUAL-SHAKER SINE VIBRATION CONTROL. Yaoqi FENG 1, Hanping QIU 1. China Academy of Space Technology (CAST) yaoqi.feng@yahoo.
ICSV4 Carns Australa 9- July, 007 RESEARCH ON DUAL-SHAKER SINE VIBRATION CONTROL Yaoq FENG, Hanpng QIU Dynamc Test Laboratory, BISEE Chna Academy of Space Technology (CAST) [email protected] Abstract
How To Understand The Results Of The German Meris Cloud And Water Vapour Product
Ttel: Project: Doc. No.: MERIS level 3 cloud and water vapour products MAPP MAPP-ATBD-ClWVL3 Issue: 1 Revson: 0 Date: 9.12.1998 Functon Name Organsaton Sgnature Date Author: Bennartz FUB Preusker FUB Schüller
Comparison of Control Strategies for Shunt Active Power Filter under Different Load Conditions
Comparson of Control Strateges for Shunt Actve Power Flter under Dfferent Load Condtons Sanjay C. Patel 1, Tushar A. Patel 2 Lecturer, Electrcal Department, Government Polytechnc, alsad, Gujarat, Inda
Calculating the high frequency transmission line parameters of power cables
< ' Calculatng the hgh frequency transmsson lne parameters of power cables Authors: Dr. John Dcknson, Laboratory Servces Manager, N 0 RW E B Communcatons Mr. Peter J. Ncholson, Project Assgnment Manager,
Single and multiple stage classifiers implementing logistic discrimination
Sngle and multple stage classfers mplementng logstc dscrmnaton Hélo Radke Bttencourt 1 Dens Alter de Olvera Moraes 2 Vctor Haertel 2 1 Pontfíca Unversdade Católca do Ro Grande do Sul - PUCRS Av. Ipranga,
The OC Curve of Attribute Acceptance Plans
The OC Curve of Attrbute Acceptance Plans The Operatng Characterstc (OC) curve descrbes the probablty of acceptng a lot as a functon of the lot s qualty. Fgure 1 shows a typcal OC Curve. 10 8 6 4 1 3 4
THE APPLICATION OF DATA MINING TECHNIQUES AND MULTIPLE CLASSIFIERS TO MARKETING DECISION
Internatonal Journal of Electronc Busness Management, Vol. 3, No. 4, pp. 30-30 (2005) 30 THE APPLICATION OF DATA MINING TECHNIQUES AND MULTIPLE CLASSIFIERS TO MARKETING DECISION Yu-Mn Chang *, Yu-Cheh
Can Auto Liability Insurance Purchases Signal Risk Attitude?
Internatonal Journal of Busness and Economcs, 2011, Vol. 10, No. 2, 159-164 Can Auto Lablty Insurance Purchases Sgnal Rsk Atttude? Chu-Shu L Department of Internatonal Busness, Asa Unversty, Tawan Sheng-Chang
Vehicle Detection and Tracking in Video from Moving Airborne Platform
Journal of Computatonal Informaton Systems 10: 12 (2014) 4965 4972 Avalable at http://www.jofcs.com Vehcle Detecton and Trackng n Vdeo from Movng Arborne Platform Lye ZHANG 1,2,, Hua WANG 3, L LI 2 1 School
DEFINING %COMPLETE IN MICROSOFT PROJECT
CelersSystems DEFINING %COMPLETE IN MICROSOFT PROJECT PREPARED BY James E Aksel, PMP, PMI-SP, MVP For Addtonal Informaton about Earned Value Management Systems and reportng, please contact: CelersSystems,
Loop Parallelization
- - Loop Parallelzaton C-52 Complaton steps: nested loops operatng on arrays, sequentell executon of teraton space DECLARE B[..,..+] FOR I :=.. FOR J :=.. I B[I,J] := B[I-,J]+B[I-,J-] ED FOR ED FOR analyze
CS 2750 Machine Learning. Lecture 3. Density estimation. CS 2750 Machine Learning. Announcements
Lecture 3 Densty estmaton Mlos Hauskrecht [email protected] 5329 Sennott Square Next lecture: Matlab tutoral Announcements Rules for attendng the class: Regstered for credt Regstered for audt (only f there
An Alternative Way to Measure Private Equity Performance
An Alternatve Way to Measure Prvate Equty Performance Peter Todd Parlux Investment Technology LLC Summary Internal Rate of Return (IRR) s probably the most common way to measure the performance of prvate
Bayesian Network Based Causal Relationship Identification and Funding Success Prediction in P2P Lending
Proceedngs of 2012 4th Internatonal Conference on Machne Learnng and Computng IPCSIT vol. 25 (2012) (2012) IACSIT Press, Sngapore Bayesan Network Based Causal Relatonshp Identfcaton and Fundng Success
A Crossplatform ECG Compression Library for Mobile HealthCare Services
A Crossplatform ECG Compresson Lbrary for Moble HealthCare Servces Alexander Borodn, Yulya Zavyalova Department of Computer Scence Petrozavodsk State Unversty Petrozavodsk, Russa {aborod, yzavyalo}@cs.petrsu.ru
Project Networks With Mixed-Time Constraints
Project Networs Wth Mxed-Tme Constrants L Caccetta and B Wattananon Western Australan Centre of Excellence n Industral Optmsaton (WACEIO) Curtn Unversty of Technology GPO Box U1987 Perth Western Australa
Improved SVM in Cloud Computing Information Mining
Internatonal Journal of Grd Dstrbuton Computng Vol.8, No.1 (015), pp.33-40 http://dx.do.org/10.1457/jgdc.015.8.1.04 Improved n Cloud Computng Informaton Mnng Lvshuhong (ZhengDe polytechnc college JangSu
+ + + - - This circuit than can be reduced to a planar circuit
MeshCurrent Method The meshcurrent s analog of the nodeoltage method. We sole for a new set of arables, mesh currents, that automatcally satsfy KCLs. As such, meshcurrent method reduces crcut soluton to
8.5 UNITARY AND HERMITIAN MATRICES. The conjugate transpose of a complex matrix A, denoted by A*, is given by
6 CHAPTER 8 COMPLEX VECTOR SPACES 5. Fnd the kernel of the lnear transformaton gven n Exercse 5. In Exercses 55 and 56, fnd the mage of v, for the ndcated composton, where and are gven by the followng
Damage detection in composite laminates using coin-tap method
Damage detecton n composte lamnates usng con-tap method S.J. Km Korea Aerospace Research Insttute, 45 Eoeun-Dong, Youseong-Gu, 35-333 Daejeon, Republc of Korea [email protected] 45 The con-tap test has the
Logical Development Of Vogel s Approximation Method (LD-VAM): An Approach To Find Basic Feasible Solution Of Transportation Problem
INTERNATIONAL JOURNAL OF SCIENTIFIC & TECHNOLOGY RESEARCH VOLUME, ISSUE, FEBRUARY ISSN 77-866 Logcal Development Of Vogel s Approxmaton Method (LD- An Approach To Fnd Basc Feasble Soluton Of Transportaton
A Simple Approach to Clustering in Excel
A Smple Approach to Clusterng n Excel Aravnd H Center for Computatonal Engneerng and Networng Amrta Vshwa Vdyapeetham, Combatore, Inda C Rajgopal Center for Computatonal Engneerng and Networng Amrta Vshwa
"Research Note" APPLICATION OF CHARGE SIMULATION METHOD TO ELECTRIC FIELD CALCULATION IN THE POWER CABLES *
Iranan Journal of Scence & Technology, Transacton B, Engneerng, ol. 30, No. B6, 789-794 rnted n The Islamc Republc of Iran, 006 Shraz Unversty "Research Note" ALICATION OF CHARGE SIMULATION METHOD TO ELECTRIC
Calculation of Sampling Weights
Perre Foy Statstcs Canada 4 Calculaton of Samplng Weghts 4.1 OVERVIEW The basc sample desgn used n TIMSS Populatons 1 and 2 was a two-stage stratfed cluster desgn. 1 The frst stage conssted of a sample
Design and Development of a Security Evaluation Platform Based on International Standards
Internatonal Journal of Informatcs Socety, VOL.5, NO.2 (203) 7-80 7 Desgn and Development of a Securty Evaluaton Platform Based on Internatonal Standards Yuj Takahash and Yoshm Teshgawara Graduate School
Software project management with GAs
Informaton Scences 177 (27) 238 241 www.elsever.com/locate/ns Software project management wth GAs Enrque Alba *, J. Francsco Chcano Unversty of Málaga, Grupo GISUM, Departamento de Lenguajes y Cencas de
Efficient Project Portfolio as a tool for Enterprise Risk Management
Effcent Proect Portfolo as a tool for Enterprse Rsk Management Valentn O. Nkonov Ural State Techncal Unversty Growth Traectory Consultng Company January 5, 27 Effcent Proect Portfolo as a tool for Enterprse
Performance Analysis of Energy Consumption of Smartphone Running Mobile Hotspot Application
Internatonal Journal of mart Grd and lean Energy Performance Analyss of Energy onsumpton of martphone Runnng Moble Hotspot Applcaton Yun on hung a chool of Electronc Engneerng, oongsl Unversty, 511 angdo-dong,
Investigation of Normalization Techniques and Their Impact on a Recognition Rate in Handwritten Numeral Recognition
S C H E D A E I N F O R M A T I C A E VOLUME 19 010 Investgaton of Normalzaton Technques and Ther Impact on a Recognton Rate n Handwrtten Numeral Recognton WIESŁAW CHMIELNICKI 1, KATARZYNA STĄPOR 1 Faculty
FREQUENCY OF OCCURRENCE OF CERTAIN CHEMICAL CLASSES OF GSR FROM VARIOUS AMMUNITION TYPES
FREQUENCY OF OCCURRENCE OF CERTAIN CHEMICAL CLASSES OF GSR FROM VARIOUS AMMUNITION TYPES Zuzanna BRO EK-MUCHA, Grzegorz ZADORA, 2 Insttute of Forensc Research, Cracow, Poland 2 Faculty of Chemstry, Jagellonan
INVESTIGATION OF VEHICULAR USERS FAIRNESS IN CDMA-HDR NETWORKS
21 22 September 2007, BULGARIA 119 Proceedngs of the Internatonal Conference on Informaton Technologes (InfoTech-2007) 21 st 22 nd September 2007, Bulgara vol. 2 INVESTIGATION OF VEHICULAR USERS FAIRNESS
To manage leave, meeting institutional requirements and treating individual staff members fairly and consistently.
Corporate Polces & Procedures Human Resources - Document CPP216 Leave Management Frst Produced: Current Verson: Past Revsons: Revew Cycle: Apples From: 09/09/09 26/10/12 09/09/09 3 years Immedately Authorsaton:
行 政 院 國 家 科 學 委 員 會 補 助 專 題 研 究 計 畫 成 果 報 告 期 中 進 度 報 告
行 政 院 國 家 科 學 委 員 會 補 助 專 題 研 究 計 畫 成 果 報 告 期 中 進 度 報 告 畫 類 別 : 個 別 型 計 畫 半 導 體 產 業 大 型 廠 房 之 設 施 規 劃 計 畫 編 號 :NSC 96-2628-E-009-026-MY3 執 行 期 間 : 2007 年 8 月 1 日 至 2010 年 7 月 31 日 計 畫 主 持 人 : 巫 木 誠 共 同
ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING
ANALYZING THE RELATIONSHIPS BETWEEN QUALITY, TIME, AND COST IN PROJECT MANAGEMENT DECISION MAKING Matthew J. Lberatore, Department of Management and Operatons, Vllanova Unversty, Vllanova, PA 19085, 610-519-4390,
The Network flow Motoring System based on Particle Swarm Optimized
The Network flow Motorng System based on Partcle Swarm Optmzed Neural Network Adult Educaton College, Hebe Unversty of Archtecture, Zhangjakou Hebe 075000, Chna Abstract The compatblty of the commercal
Course outline. Financial Time Series Analysis. Overview. Data analysis. Predictive signal. Trading strategy
Fnancal Tme Seres Analyss Patrck McSharry [email protected] www.mcsharry.net Trnty Term 2014 Mathematcal Insttute Unversty of Oxford Course outlne 1. Data analyss, probablty, correlatons, vsualsaton
On the Optimal Control of a Cascade of Hydro-Electric Power Stations
On the Optmal Control of a Cascade of Hydro-Electrc Power Statons M.C.M. Guedes a, A.F. Rbero a, G.V. Smrnov b and S. Vlela c a Department of Mathematcs, School of Scences, Unversty of Porto, Portugal;
Modelling of Web Domain Visits by Radial Basis Function Neural Networks and Support Vector Machine Regression
Modellng of Web Doman Vsts by Radal Bass Functon Neural Networks and Support Vector Machne Regresson Vladmír Olej, Jana Flpová Insttute of System Engneerng and Informatcs Faculty of Economcs and Admnstraton,
Linear Circuits Analysis. Superposition, Thevenin /Norton Equivalent circuits
Lnear Crcuts Analyss. Superposton, Theenn /Norton Equalent crcuts So far we hae explored tmendependent (resste) elements that are also lnear. A tmendependent elements s one for whch we can plot an / cure.
A Hierarchical Anomaly Network Intrusion Detection System using Neural Network Classification
IDC IDC A Herarchcal Anomaly Network Intruson Detecton System usng Neural Network Classfcaton ZHENG ZHANG, JUN LI, C. N. MANIKOPOULOS, JAY JORGENSON and JOSE UCLES ECE Department, New Jersey Inst. of Tech.,
Conversion between the vector and raster data structures using Fuzzy Geographical Entities
Converson between the vector and raster data structures usng Fuzzy Geographcal Enttes Cdála Fonte Department of Mathematcs Faculty of Scences and Technology Unversty of Combra, Apartado 38, 3 454 Combra,
IMPACT ANALYSIS OF A CELLULAR PHONE
4 th ASA & μeta Internatonal Conference IMPACT AALYSIS OF A CELLULAR PHOE We Lu, 2 Hongy L Bejng FEAonlne Engneerng Co.,Ltd. Bejng, Chna ABSTRACT Drop test smulaton plays an mportant role n nvestgatng
where the coordinates are related to those in the old frame as follows.
Chapter 2 - Cartesan Vectors and Tensors: Ther Algebra Defnton of a vector Examples of vectors Scalar multplcaton Addton of vectors coplanar vectors Unt vectors A bass of non-coplanar vectors Scalar product
A Multi-mode Image Tracking System Based on Distributed Fusion
A Mult-mode Image Tracng System Based on Dstrbuted Fuson Ln zheng Chongzhao Han Dongguang Zuo Hongsen Yan School of Electroncs & nformaton engneerng, X an Jaotong Unversty X an, Shaanx, Chna [email protected]
VEHICLE DETECTION BY USING REAR PARTS AND TRACKING SYSTEM
IJRET: Internatonal Journal of Research n Engneerng and Technology eissn: 2319-1163 pissn: 2321-7308 VEHICLE DETECTION BY USING REAR PARTS AND TRACKING SYSTEM Yogn Ashokrao Kanhegaonkar 1, Jagtap Rupal
Minimal Coding Network With Combinatorial Structure For Instantaneous Recovery From Edge Failures
Mnmal Codng Network Wth Combnatoral Structure For Instantaneous Recovery From Edge Falures Ashly Joseph 1, Mr.M.Sadsh Sendl 2, Dr.S.Karthk 3 1 Fnal Year ME CSE Student Department of Computer Scence Engneerng
MATHEMATICAL ENGINEERING TECHNICAL REPORTS. Sequential Optimizing Investing Strategy with Neural Networks
MATHEMATICAL ENGINEERING TECHNICAL REPORTS Sequental Optmzng Investng Strategy wth Neural Networks Ryo ADACHI and Akmch TAKEMURA METR 2010 03 February 2010 DEPARTMENT OF MATHEMATICAL INFORMATICS GRADUATE
IDENTIFICATION AND CORRECTION OF A COMMON ERROR IN GENERAL ANNUITY CALCULATIONS
IDENTIFICATION AND CORRECTION OF A COMMON ERROR IN GENERAL ANNUITY CALCULATIONS Chrs Deeley* Last revsed: September 22, 200 * Chrs Deeley s a Senor Lecturer n the School of Accountng, Charles Sturt Unversty,
Descriptive Models. Cluster Analysis. Example. General Applications of Clustering. Examples of Clustering Applications
CMSC828G Prncples of Data Mnng Lecture #9 Today s Readng: HMS, chapter 9 Today s Lecture: Descrptve Modelng Clusterng Algorthms Descrptve Models model presents the man features of the data, a global summary
PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 12
14 The Ch-squared dstrbuton PSYCHOLOGICAL RESEARCH (PYC 304-C) Lecture 1 If a normal varable X, havng mean µ and varance σ, s standardsed, the new varable Z has a mean 0 and varance 1. When ths standardsed
Intelligent Voice-Based Door Access Control System Using Adaptive-Network-based Fuzzy Inference Systems (ANFIS) for Building Security
Journal of Computer Scence 3 (5): 274-280, 2007 ISSN 1549-3636 2007 Scence Publcatons Intellgent Voce-Based Door Access Control System Usng Adaptve-Network-based Fuzzy Inference Systems (ANFIS) for Buldng
Mining Multiple Large Data Sources
The Internatonal Arab Journal of Informaton Technology, Vol. 7, No. 3, July 2 24 Mnng Multple Large Data Sources Anmesh Adhkar, Pralhad Ramachandrarao 2, Bhanu Prasad 3, and Jhml Adhkar 4 Department of
Chapter 4 ECONOMIC DISPATCH AND UNIT COMMITMENT
Chapter 4 ECOOMIC DISATCH AD UIT COMMITMET ITRODUCTIO A power system has several power plants. Each power plant has several generatng unts. At any pont of tme, the total load n the system s met by the
A FEATURE SELECTION AGENT-BASED IDS
A FEATURE SELECTION AGENT-BASED IDS Emlo Corchado, Álvaro Herrero and José Manuel Sáz Department of Cvl Engneerng, Unversty of Burgos C/Francsco de Vtora s/n., 09006, Burgos, Span Phone: +34 947259395,
CHOLESTEROL REFERENCE METHOD LABORATORY NETWORK. Sample Stability Protocol
CHOLESTEROL REFERENCE METHOD LABORATORY NETWORK Sample Stablty Protocol Background The Cholesterol Reference Method Laboratory Network (CRMLN) developed certfcaton protocols for total cholesterol, HDL
Logistic Regression. Lecture 4: More classifiers and classes. Logistic regression. Adaboost. Optimization. Multiple class classification
Lecture 4: More classfers and classes C4B Machne Learnng Hlary 20 A. Zsserman Logstc regresson Loss functons revsted Adaboost Loss functons revsted Optmzaton Multple class classfcaton Logstc Regresson
IT09 - Identity Management Policy
IT09 - Identty Management Polcy Introducton 1 The Unersty needs to manage dentty accounts for all users of the Unersty s electronc systems and ensure that users hae an approprate leel of access to these
Open Access A Load Balancing Strategy with Bandwidth Constraint in Cloud Computing. Jing Deng 1,*, Ping Guo 2, Qi Li 3, Haizhu Chen 1
Send Orders for Reprnts to [email protected] The Open Cybernetcs & Systemcs Journal, 2014, 8, 115-121 115 Open Access A Load Balancng Strategy wth Bandwdth Constrant n Cloud Computng Jng Deng 1,*,
Institute of Informatics, Faculty of Business and Management, Brno University of Technology,Czech Republic
Lagrange Multplers as Quanttatve Indcators n Economcs Ivan Mezník Insttute of Informatcs, Faculty of Busness and Management, Brno Unversty of TechnologCzech Republc Abstract The quanttatve role of Lagrange
An artificial Neural Network approach to monitor and diagnose multi-attribute quality control processes. S. T. A. Niaki*
Journal of Industral Engneerng Internatonal July 008, Vol. 4, No. 7, 04 Islamc Azad Unversty, South Tehran Branch An artfcal Neural Network approach to montor and dagnose multattrbute qualty control processes
Study on Model of Risks Assessment of Standard Operation in Rural Power Network
Study on Model of Rsks Assessment of Standard Operaton n Rural Power Network Qngj L 1, Tao Yang 2 1 Qngj L, College of Informaton and Electrcal Engneerng, Shenyang Agrculture Unversty, Shenyang 110866,
A study on the ability of Support Vector Regression and Neural Networks to Forecast Basic Time Series Patterns
A study on the ablty of Support Vector Regresson and Neural Networks to Forecast Basc Tme Seres Patterns Sven F. Crone, Jose Guajardo 2, and Rchard Weber 2 Lancaster Unversty, Department of Management
Implementation of Deutsch's Algorithm Using Mathcad
Implementaton of Deutsch's Algorthm Usng Mathcad Frank Roux The followng s a Mathcad mplementaton of Davd Deutsch's quantum computer prototype as presented on pages - n "Machnes, Logc and Quantum Physcs"
ECE544NA Final Project: Robust Machine Learning Hardware via Classifier Ensemble
1 ECE544NA Fnal Project: Robust Machne Learnng Hardware va Classfer Ensemble Sa Zhang, [email protected] Dept. of Electr. & Comput. Eng., Unv. of Illnos at Urbana-Champagn, Urbana, IL, USA Abstract In
RELIABILITY, RISK AND AVAILABILITY ANLYSIS OF A CONTAINER GANTRY CRANE ABSTRACT
Kolowrock Krzysztof Joanna oszynska MODELLING ENVIRONMENT AND INFRATRUCTURE INFLUENCE ON RELIABILITY AND OPERATION RT&A # () (Vol.) March RELIABILITY RIK AND AVAILABILITY ANLYI OF A CONTAINER GANTRY CRANE
A Comparative Study of Data Clustering Techniques
A COMPARATIVE STUDY OF DATA CLUSTERING TECHNIQUES A Comparatve Study of Data Clusterng Technques Khaled Hammouda Prof. Fakhreddne Karray Unversty of Waterloo, Ontaro, Canada Abstract Data clusterng s a
Performance Analysis and Coding Strategy of ECOC SVMs
Internatonal Journal of Grd and Dstrbuted Computng Vol.7, No. (04), pp.67-76 http://dx.do.org/0.457/jgdc.04.7..07 Performance Analyss and Codng Strategy of ECOC SVMs Zhgang Yan, and Yuanxuan Yang, School
