KOHONEN'S SELF ORGANIZING NETWORKS WITH "CONSCIENCE"

Size: px
Start display at page:

Download "KOHONEN'S SELF ORGANIZING NETWORKS WITH "CONSCIENCE""

Transcription

1 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 1 KOHONEN'S SELF ORGANIZING NETWORKS WITH "CONSCIENCE" By: Dr. M. Turhan (Tury) Taner, Rock Sold Images November 1997 INTRODUCTION: The effectveness of the nterpretve use "Sesmc Attrbutes" depends on the dscrmnaton ablty of the chosen set of attrbutes. As n many mathematcal problems, some attrbutes may be necessary; a few may be suffcent. Ths could be determned by expermentng wth logcal combnatons of varous mxtures of attrbutes. In cases where well logs or lthologcal columns are avalable, Feed Forward fully connected Neural Networks may be traned n a supervsed manner. In cases where no well nformaton s avalable, we wll have to use some unsupervsed methodology to cluster the data and check to see f t wll have any relaton to our experence based nterpretaton. In our case the data s the computed set of attrbutes. The user selects a set of attrbutes to be used n clusterng or n classfcaton. There are several unsupervsed clusterng methods to do the job. One of the smplest s to cluster the data sequentally. Assume that we are gven a set of data traces, where each sample s represented by a vector of attrbutes. Ths can be vewed as data represented n an N dmensonal space, where each attrbute s one of the axes. In the begnnng we compute a smple mean of all of the attrbutes; then we collect all of the data samples wthn some dstance to ths mean value. As we are checkng all of the samples, we can dentfy the data samples wthn the allowed dstance as members of the frst cluster and take them away from the further computaton. After all of the data sets are computed we have formed the frst cluster and dentfcaton. Whle ths s gong on we can compute the average of the rejected samples, whch wll represent the mean of the second set. We contnue clusterng n ths manner untl no sgnfcant data samples are left. The problem wth ths procedure, although t may be fast, s that all clusters are arbtrary and no reasonable relatonshp s mantaned between the clusters. Ths feature, as we wll dscuss later, represents an mportant aspect of Kohonen's method where clusters are formed n such a way that they relate to one another n an organzed manner. For example, clusterng for the sand, sandy-shale, shaly-sand and shale cluster may occupy adjacent locatons on the organzed map. Ths s analogous to the human brans "assocatve memory". Another, more classcal approach s the development of statstcs of the data set by formng the covarance matrx and computng the prncpal components. Ths wll gve us Egen values and Egen vectors from the most to the least mportant. The Egen vector correspondng to the most sgnfcant Egen value wll represent the attrbute combnaton whch s most abundant n the data set. The next Egen value wll represent the next most abundant and so on. Once these are determned, we can classfy the data set by computng the dstance between the data set and the Egen vectors and selectng the one wth the mnmum Eucldean dstance. Ths method wll form clusters based on populaton and may not have any relatonshp to adjacent Egen values and vectors. Furthermore, we wll have to solve a very large matrx, composed of correlatons between all pars of the data set. KOHONEN'S ANALOGY: Kohonen's self-organzng maps (SOM) are smple analogs of the human bran s way of organzng nformaton n a logcal manner. Research has shown that the cerebral cortex of the human bran s dvded nto functonal subdvsons. The cortex contans bllons of neurons wth many bllons of connectons (synapses) between them. The subdvson of the cortex s an orderly manner. The most mportant ones are motor cortex, sensory cortex, vsual cortex and audtory cortex. For example, the audtory cortex s also subdvded nto many cells, each functonng as cogntve parts for some audtory sgnal. It s theorzed that some of these are traned n a supervsed manner, and some are developed n an unsupervsed self-organzng manner. Topologcally adjacent areas perform somewhat related cogntve functons. Kohonen's method emulates the unsupervsed learnng n an elegant, extremely smple manner. Durng the self-organzng procedure the topologcally close relatonshp of the organzed nformaton s mantaned. Intally, a large area s treated n a smlar fashon. Later n the teraton ths zone shrnks. A number of other computatonal procedures were ntroduced after the orgnal Kohonen algorthm. Snce "wnner take all" logc

2 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: may result n one neuron domnatng the tranng, some modfcaton s necessary. I wll descrbe the orgnal method wth only one smple modfcaton, called the " conscence. Kohonen's method conssts of one layer of neurons and uses the method of compettve learnng wth "wnner take all ". The modfcaton looks at the wnnng rate of each neuron, and f t sees any neuron domnatng, t uses the "conscence" algorthm and lets other neurons have a chance n the learnng procedure. METHOD: As n the case of covarance studes, nput to the self-organzng maps wll have to be normalzed. That s each set of attrbutes has to be scaled so that ther RMS value wll be equal to unty. Ths wll prevent any attrbute arbtrarly domnatng the clusterng. Kohonen's self-organzng maps consst of one layer of neurons organzed n one, two and mult-dmensonal arrays. Each neuron has as many nput connectons as there are number of attrbutes to be used n the classfcaton, as shown on fgure 1. The tranng procedure conssts of fndng the neuron wth weghts closest to the nput data vector and declarng that neuron as the wnnng neuron. Then the weghts of all of the neurons n the vcnty of the wnnng neuron are adjusted, by an amount nversely proportonal to the dstance. The radus of the accepted vcnty s reduced as the teraton numbers ncrease. The tranng process s termnated f RMS errors of all of the nput are reduced to an acceptable level or a prescrbed number of teratons are reached. There are two methods to determne the smlarty. In the frst method, each nput s weghted by a neuron s correspondng weght vector and the results are summed. Ths represents the net nput of the partcular neuron. Let k represent the k'th neuron and N attrbutes are used, then the net nput wll be (n terms of a vector scalar product); net k = N = 1 [ x( ). w(, k)] (1) The vector scalar product (1) wll gve the projecton of one vector on to the other. If the unt vectors are defned n the x() and w() drectons, ther dot product yelds the cosne of the angle between x and w vectors. Ths s one of the reasons we normalze each set of attrbutes to have the zero mean and ther RMS ampltude be unty. A net nput of 1 would represent two collnear vectors; they pont n the same drecton and ther parameters are smlar. A value of 0 wll mean that two vectors are perpendcular to each other, thus they are not smlar. The second method to measure the smlarty of two vectors s to compute the Eucldean dstance between two vectors, as; net k = N = 1 [ x( ) w(, k)] () In ths case a net result of zero wll mean that the two vectors are dentcal. A value of close to twce to ther normalzed ampltudes wll mean that they are n opposte drectons, thus they are not smlar. These are repeated for all of the neurons and the neuron wth largest output of dot product or the mnmum Eucldean dstance s chosen as the wnner. In order to mantan smlarty of topologcally close neurons, weghts of all neurons wthn a selected radus are adjusted. The wnnng neuron s adjusted by the most amount, whch wll brng the weghts of the wnnng neuron closer to the nput data values. All other neurons are adjusted by lesser amounts, nversely proportonal to ther dstance from the wnnng neuron. As the teraton progresses and RMS error reduces, the radus of correcton s also gradually reduced. Ths wll eventually become one neuron dstance; thus no other neuron s adjusted other than the wnnng neuron. In the case where one neuron s contnually the wnnng neuron, then ts computed dstance s also modfed by some amount to allow other neurons wn. Ths process s called the "conscence". At the concluson of tranng, each neuron's weght represent reference data, whch could later be used for classfcaton. If the nput s sesmc attrbutes, then the weghts represent the reference attrbutes (.e. the mean) of each classfcaton (.e. cluster). The classfcaton process at the concluson of tranng s then, the computaton of each nput data sets dstance to each of the neurons and classfy the nput as belongng to the class represented by the wnnng neuron.

3 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 3 Fgure 1. Connecton of nput data to a Neural Node TOPOLOGICAL ORGANIZATION: Neurons can be organzed n any topologcal manner. Fgure shows a one-dmensonal organzaton. Ths wll allow us clusterng wth a one-dmensonal topologcal relatonshp. That s, each adjacent neuron wll have smaller dfferences than the neurons two spaces away. The dfference wll gradually ncrease wth ncreasng dstance. Each neuron s connected to the nput data wth ther own weghts. In the case of the SOM these weghts wll be equvalent to the normalzed attrbutes representng the centrod coordnates of each cluster. Ths feature of the SOM s very dfferent than the feed forward fully connected Artfcal Neural Networks, where the weghts do not relate drectly to the nput data. Fgure Input to One Dmensonal Neural Topology

4 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 4 The same characterstcs wll preval n two-dmensonal mappng as shown on fgure 3. Ths tme we have four adjacent neurons to each neuron, east, west, north and south drecton. The dfferences wll be nversely proportonal to the topologcal dstance between the neurons. We can thnk of three dmensonal mappng the same way. In practce, one or two-dmensonal mappng s usually used. I wll dscuss the case of two-dmensonal mappng n sesmc nterpretaton. The mappng confguraton wll affect only the defnton of the neghborhood functon; thus t s very easy to handle the varous dmensonalty of the maps. USES OF SELF-ORGANIZING MAPS: SOM's can be used n sesmc nterpretaton n two prncpal ways, one as a cluster generator and the other as a classfer. Input data conssts of -D or 3-D trace segments or ther attrbutes. Each data sample and correspondng attrbutes representng a pont n the N dmensonal space s used to form a set of weghts for each Neural node accordng to the network topology. These weghts are called the assocated memory, because they represent reference attrbutes of each cluster and they have some organzed assocaton to ther neghbors. Ths process takes a large number of teratons, of the order of tens of thousands. Measurng the RMS error reducton after each teraton can montor convergence. Once the convergence reaches a satsfactory mnmum, then the weghts are saved for the classfcaton. Snce each weght vector represents a reference attrbute of a partcular cluster, we perform classfcaton by computng the Eucldean dstance between each data sample and the weght vector of each node, and select the node wth mnmum dstance. The SOM s an unsupervsed learnng algorthm; t does not gve us drect nformaton on the actual physcal or lthologcal classfcaton of the data samples. It merely clusters and classfes the data set based on the set of attrbutes used. We wll need to attach lthologcal meanng after the classfcaton s completed by the use of nearby well nformaton or based on our experence n the area. If there s a lthologcal column avalable n the general vcnty of the sesmc data, then we can experment wth dfferent sets of attrbutes to see whch set gves us smlar classfcaton boundares as the lthologcal column. Fgure 3. Input to -Dmensonal Neural Topology

5 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 5 COMPUTATIONAL PROCEDURE: I wll outlne the computatonal procedure step by step ncludng some dscusson assocated wth each step. 1. Data Intalzaton: Once the composton of the attrbutes s selected, each attrbute set wll be normalzed (scaled ) to RMS=1. Ths wll prevent one attrbute from overpowerng n clusterng and classfcaton. If, however, we consder some attrbutes more dagnostc than the others, we can upward adjust ther RMS ampltudes. Ths wll make ths partcular attrbute's dstance be more promnent than the rest, hence mnmzaton wll be affected by t.. Intalzaton of Neural Weghts: Each neuron wll have one weght for each attrbute. To begn, we assgn small random numbers to each of the weghts, as n the case of other Neural Network computatons. In the begnnng we wll also ntalze the neghborhood radus. Ths radus wll be large at the begnnng to cover a larger number of adjacent nodes. As the teraton proceeds, ths radus may be systematcally reduced to a unt dstance. We wll also ntalze a set of counters that wll ndcate the number of consecutve tmes a Neural node s declared as the wnner. In the "conscence" algorthm, f a node wns excessvely, t wll feel bad and allows others to wn. Ths s accomplshed by artfcally reducng the computed weghts, causng other nodes to wn. 3. Selecton of Input data: Select nput data at random from the gven set of sesmc data traces and ther attrbutes. Ths wll prevent smlar neghborng attrbutes from over nfluencng the weght-updatng scheme. All of the data samples are nput at each teraton cycle. 4. Computaton of Eucldean Dstance or Vector Dot Product and determnng the wnnng Node: The Eucldean dstance or Vector dot product s computed between the nput data set (vector) and each of the Neural Nodes ( as shown on equatons 1 and ). The Node wth the mnmum Eucldean dstance or maxmum vector dot product s declared the wnner. Ths determnes the node wth the most smlar set of weghts to the nput data attrbutes. Before the wnnng node s determned, the number of consecutve wns s checked. If the node had exceeded the lmt, then the results are reduced proportonally to allow another node to wn. The losng node counters are reset to zero, and the wnnng node counter s ncremented by one. In some nstances one node may contnually wn, thus overpowerng all the other nodes. To prevent ths DeSeno (1988) has ntroduced an algorthm called the "conscence". Ths algorthm keeps track of the wnnng neurons and f they exceed some level of number of wnnng, ther Eucldean dstances are ncreased or the dot products are decreased to allow the other neurons a chance to wn. Let the output of wnnng 'th element gven by; y =1 f w X < wj X for all j values, except, j y = 0 otherwse. (3) A bas s developed for each neuron based on the number of tmes t has won the competton shown on eq. (3) Let p represent the fracton of tme a neuron wns, then we can update the bas factor by;. p new old old = p + β [ y p ], (4) where 0 < β <<1. We are usng = β.. We compute now, the conscence factor; b = C 1/ N p ), (5) (

6 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 6 where C s the bas factor, and N s the number of neurons. Some of the examples on varous book use C=10. We have to experment wth ths value. Once the conscence factor s determned, new ( conscence modfed ) Eucldean dstances wll be computed as; z =1 f w X b < wj X b j for all j values, except, j z = 0 otherwse. (6) Ths wll determne the new wnner. We wll contnue the process by makng the weght vector of the wnnng neuron closer the nput data, as descrbed below. 5. Updatng the weghts of wnnng node and ts neghbors: The wnnng nodes weghts wll be modfed by ; w( n 1) ( k, = wn ( k, + ( n)[ x( wn ( k, ] + η (7) where n s the number of the teraton, k s the wnnng node and j s the j'th attrbute. η(n) s the modfcaton weght as a decreasng functon of the number of teratons. Ths s also called the learnng rate. All of the nodes wthn the assocatve radus wll also be modfed, nversely proportonal to the dstance on the map. If we assgn one, two or more dmensonalty to each node, we can compute the relatve dstance from the wnnng node to each other node and can easly compute the modfcatons. Let the radus d(m,k,n) be represented as a decreasng functon of the teraton number. 6. Update neghborng nodes accordng to ther dstances to the wnnng node; w ( ( n 1) m, = wn ( m, + d( m, k, n). ( n)[ x( wn ( m, ] + η (8) where d(m,k,n) s a functon of teraton number n, and the dstance between wnnng node k and ts neghbor m. All other nodes outsde the assocatve radus wll have no correcton. 7. Update assocatve correcton radus; In the begnnng the ntal radus may be kept long enough to cover most of the neural nodes. Ths s reduced as the teraton number ncreases. d(m,k,n) can be made any functon nversely proportonal to the teraton number n, and the dstance between the wnnng node k and the node to be adjusted m. 8. Check convergence rate to contnue or to go to fnsh teraton: We compute RMS error (square root of the sum of error squares, whch s the square of the dstance between the nput and the wnnng neural nodes weghts dvded by the number of nputs) and check f t s below some prescrbed level. If t s, then we have developed the desred weghts, whch can be used for classfcaton. If not, then we go back for another set of teraton (back to step 3). 9. Qualty control: One of the man characterstcs of Kohonen's mappng s the determnaton neurons wth topologcally assocatve weght sets. Therefore, a good qualty check s to see f the Eucldean dstance between neurons computed from ther weghts do satsfy the orgnal topologcal desgn. That s, computed dstances between neural nodes should be proportonal to the topologcal dstances of the desred map. In other words, fnal map coordnates should resemble shape of the user specfed map. If ths condton s not satsfed, some addtonal teraton may be needed. 10. Save computed weghts as classfers or as reference attrbutes: The computed weghts of the neural nodes are the memory functons, whch are used for classfcaton. They should have smaller dfferences between adjacent neurons. In our case they represent the Attrbute values of

7 Kohonen's Self Organzng Maps and ther use n Interpretaton, Dr. M. Turhan (Tury) Taner, Rock Sold Images Page: 7 each classfcaton. (For actual values we have to remove the scalng we appled to the attrbutes durng normalzng.) 11. Use computed reference attrbutes to classfy nput data set: In the classfcaton stage for each nput, we compute the Eucldean dstance to each neural node and classfy the nput as a member of the class represented by the wnnng node. Ths s very smlar to the actual teraton, except we no longer adjust the weghts. 1. Valdty check of the results: The "conscence" condton tres to develop a unform or equally balanced clusterng. Therefore we should expect that each classfcaton should have about the same number of members and smlar varance. Durng the classfcaton process, the number of members of each class and ther statstcs are accumulated, whch could then be used for a check of valdty. These results, as well as other qualty control statstcs, are output to the user for proper control of the process.. CONCLUSIONS: Kohonen's Self-Organzng Map method can be used n several dfferent ways for lthologc nterpretaton. One of the most obvous ways s to use a typcal porton of a -D or 3-D data set to tran the network; then use the reference weghts as classfers for the rest of the data set. If we have a lthologcal column, we can experment wth a number of dfferent and logcal attrbute combnatons to see whch clusters are most smlar to the lthologcal column. In ths elegant method, there s no restrcton to the dmenson of the nput. It could be one, two or many dmensonal nput vectors, defnng varous physcal attrbutes or geometrcal condtons. Ths means that, we may be able to use the SOM for vsual type classfcatons. Ths we wll have to experment wth n the near future. ACKNOWLEDMENTS: I would lke to express my thanks to James Schuelke and John Quren of Mobl R&D Corporaton for ther nvaluable help and drecton. To them and to Graham Jago goes many thanks for gentle edtng of the text. The value of the SOM technque n nterpretaton became more obvous n our conversatons. REFERENCES: Proceedngs of the IEEE, 1990, Specal ssue on Neural Networks; 1. Neural Networks, theory and modelng (September ssue).neural networks, analyss, technques and applcatons (October ssue) (These ssues have extensve lterature and background artcles) Bharath, R, and Drosen, J, 1994, "Neural Network Computng", Publshed b Wndcrest/McGraw-Hll, New York (Ths book contans examples of some prmtve functons used n Neural computaton. C-code programs are ncluded on a dskette) Chester, M, 1993, "Neural networks, A Tutoral", Publshed by PTR Prentce Hall, Englewood Clffs, New Jersey. (Ths book s recommended to the begnners of the Neural Networks) DeSeno, D., "Addng a Conscence to compettve learnng." IEEE Internatonal Conference on Neural Networks, Vol. 1, pp , San Dego CA. Haykn, S. 1994, "Neural Networks, A Comprehensve Foundaton", Publshed by Macmllan College Publshng Company, New York. (Ths book contans one of the most comprehensve dscussons of all types of Neural Networks ncludng the Self-Organzed Maps). Kohonen, T., 1988, "Self-Organzng and Assocatve Memory", 3 rd. ed., Publshed by Sprnger Verlag, New York. Pao Y. H., 1989, "Adaptve Pattern Recognton and Neural Networks", Addson-Wesley Publshng Company.

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

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points;

Subspace clustering. Clustering. Fundamental to all clustering techniques is the choice of distance measure between data points; Subspace clusterng Clusterng Fundamental to all clusterng technques s the choce of dstance measure between data ponts; D q ( ) ( ) 2 x x = x x, j k = 1 k jk Squared Eucldean dstance Assumpton: All features

More information

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

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

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

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

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

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

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

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

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

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Why consder unlabeled samples?. Collectng and labelng large set of samples s costly Gettng recorded speech s free, labelng s tme consumng 2. Classfer could be desgned

More information

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

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

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

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

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

Smoothing Spline ANOVA for variable screening

Smoothing Spline ANOVA for variable screening Smoothng Splne ANOVA for varable screenng a useful tool for metamodels tranng and mult-objectve optmzaton L. Rcco, E. Rgon, A. Turco Outlne RSM Introducton Possble couplng Test case MOO MOO wth Game Theory

More information

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

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning

Outline. Type of Machine Learning. Examples of Application. Unsupervised Learning Outlne Artfcal Intellgence and ts applcatons Lecture 8 Unsupervsed Learnng Professor Danel Yeung danyeung@eee.org Dr. Patrck Chan patrckchan@eee.org South Chna Unversty of Technology, Chna Introducton

More information

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

Programming in Fortran 90 : 2017/2018

Programming in Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Programmng n Fortran 90 : 2017/2018 Exercse 1 : Evaluaton of functon dependng on nput Wrte a program who evaluate the functon f (x,y) for any two user specfed values

More information

Unsupervised Learning

Unsupervised Learning Pattern Recognton Lecture 8 Outlne Introducton Unsupervsed Learnng Parametrc VS Non-Parametrc Approach Mxture of Denstes Maxmum-Lkelhood Estmates Clusterng Prof. Danel Yeung School of Computer Scence and

More information

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

Data Representation in Digital Design, a Single Conversion Equation and a Formal Languages Approach

Data Representation in Digital Design, a Single Conversion Equation and a Formal Languages Approach Data Representaton n Dgtal Desgn, a Sngle Converson Equaton and a Formal Languages Approach Hassan Farhat Unversty of Nebraska at Omaha Abstract- In the study of data representaton n dgtal desgn and computer

More information

Module Management Tool in Software Development Organizations

Module Management Tool in Software Development Organizations Journal of Computer Scence (5): 8-, 7 ISSN 59-66 7 Scence Publcatons Management Tool n Software Development Organzatons Ahmad A. Al-Rababah and Mohammad A. Al-Rababah Faculty of IT, Al-Ahlyyah Amman Unversty,

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

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

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

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

Available online at   Available online at   Advanced in Control Engineering and Information Science Avalable onlne at wwwscencedrectcom Avalable onlne at wwwscencedrectcom Proceda Proceda Engneerng Engneerng 00 (2011) 15000 000 (2011) 1642 1646 Proceda Engneerng wwwelsevercom/locate/proceda Advanced

More information

Efficient Segmentation and Classification of Remote Sensing Image Using Local Self Similarity

Efficient Segmentation and Classification of Remote Sensing Image Using Local Self Similarity ISSN(Onlne): 2320-9801 ISSN (Prnt): 2320-9798 Internatonal Journal of Innovatve Research n Computer and Communcaton Engneerng (An ISO 3297: 2007 Certfed Organzaton) Vol.2, Specal Issue 1, March 2014 Proceedngs

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

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System

Fuzzy Modeling of the Complexity vs. Accuracy Trade-off in a Sequential Two-Stage Multi-Classifier System Fuzzy Modelng of the Complexty vs. Accuracy Trade-off n a Sequental Two-Stage Mult-Classfer System MARK LAST 1 Department of Informaton Systems Engneerng Ben-Guron Unversty of the Negev Beer-Sheva 84105

More information

Performance Assessment and Fault Diagnosis for Hydraulic Pump Based on WPT and SOM

Performance Assessment and Fault Diagnosis for Hydraulic Pump Based on WPT and SOM Performance Assessment and Fault Dagnoss for Hydraulc Pump Based on WPT and SOM Be Jkun, Lu Chen and Wang Zl PERFORMANCE ASSESSMENT AND FAULT DIAGNOSIS FOR HYDRAULIC PUMP BASED ON WPT AND SOM. Be Jkun,

More information

Backpropagation: In Search of Performance Parameters

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

More information

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

Accounting for the Use of Different Length Scale Factors in x, y and z Directions

Accounting for the Use of Different Length Scale Factors in x, y and z Directions 1 Accountng for the Use of Dfferent Length Scale Factors n x, y and z Drectons Taha Soch (taha.soch@kcl.ac.uk) Imagng Scences & Bomedcal Engneerng, Kng s College London, The Rayne Insttute, St Thomas Hosptal,

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 Clustering Algorithm for Chinese Adjectives and Nouns 1

A Clustering Algorithm for Chinese Adjectives and Nouns 1 Clusterng lgorthm for Chnese dectves and ouns Yang Wen, Chunfa Yuan, Changnng Huang 2 State Key aboratory of Intellgent Technology and System Deptartment of Computer Scence & Technology, Tsnghua Unversty,

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

VISUAL SELECTION OF SURFACE FEATURES DURING THEIR GEOMETRIC SIMULATION WITH THE HELP OF COMPUTER TECHNOLOGIES

VISUAL SELECTION OF SURFACE FEATURES DURING THEIR GEOMETRIC SIMULATION WITH THE HELP OF COMPUTER TECHNOLOGIES UbCC 2011, Volume 6, 5002981-x manuscrpts OPEN ACCES UbCC Journal ISSN 1992-8424 www.ubcc.org VISUAL SELECTION OF SURFACE FEATURES DURING THEIR GEOMETRIC SIMULATION WITH THE HELP OF COMPUTER TECHNOLOGIES

More information

High-Boost Mesh Filtering for 3-D Shape Enhancement

High-Boost Mesh Filtering for 3-D Shape Enhancement Hgh-Boost Mesh Flterng for 3-D Shape Enhancement Hrokazu Yagou Λ Alexander Belyaev y Damng We z Λ y z ; ; Shape Modelng Laboratory, Unversty of Azu, Azu-Wakamatsu 965-8580 Japan y Computer Graphcs Group,

More information

Unsupervised Neural Network Adaptive Resonance Theory 2 for Clustering

Unsupervised Neural Network Adaptive Resonance Theory 2 for Clustering 3 rd Internatonal G raduate Conference on Engneerng, Scence and Humantes (IGCESH) School of Graduate Studes Unverst Teknolog Malaysa 4 November 00 Unsupervsed Neural Network Adaptve Resonance Theory for

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

An Improved Neural Network Algorithm for Classifying the Transmission Line Faults

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

More information

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps

Visual Thesaurus for Color Image Retrieval using Self-Organizing Maps Vsual Thesaurus for Color Image Retreval usng Self-Organzng Maps Chrstopher C. Yang and Mlo K. Yp Department of System Engneerng and Engneerng Management The Chnese Unversty of Hong Kong, Hong Kong ABSTRACT

More information

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

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

More information

An Entropy-Based Approach to Integrated Information Needs Assessment

An Entropy-Based Approach to Integrated Information Needs Assessment Dstrbuton Statement A: Approved for publc release; dstrbuton s unlmted. An Entropy-Based Approach to ntegrated nformaton Needs Assessment June 8, 2004 Wllam J. Farrell Lockheed Martn Advanced Technology

More information

Mathematics 256 a course in differential equations for engineering students

Mathematics 256 a course in differential equations for engineering students Mathematcs 56 a course n dfferental equatons for engneerng students Chapter 5. More effcent methods of numercal soluton Euler s method s qute neffcent. Because the error s essentally proportonal to the

More information

Simulation Based Analysis of FAST TCP using OMNET++

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

More information

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

Object-Based Techniques for Image Retrieval

Object-Based Techniques for Image Retrieval 54 Zhang, Gao, & Luo Chapter VII Object-Based Technques for Image Retreval Y. J. Zhang, Tsnghua Unversty, Chna Y. Y. Gao, Tsnghua Unversty, Chna Y. Luo, Tsnghua Unversty, Chna ABSTRACT To overcome the

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

Hierarchical clustering for gene expression data analysis

Hierarchical clustering for gene expression data analysis Herarchcal clusterng for gene expresson data analyss Gorgo Valentn e-mal: valentn@ds.unm.t Clusterng of Mcroarray Data. Clusterng of gene expresson profles (rows) => dscovery of co-regulated and functonally

More information

CMPS 10 Introduction to Computer Science Lecture Notes

CMPS 10 Introduction to Computer Science Lecture Notes CPS 0 Introducton to Computer Scence Lecture Notes Chapter : Algorthm Desgn How should we present algorthms? Natural languages lke Englsh, Spansh, or French whch are rch n nterpretaton and meanng are not

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

K-means and Hierarchical Clustering

K-means and Hierarchical Clustering Note to other teachers and users of these sldes. Andrew would be delghted f you found ths source materal useful n gvng your own lectures. Feel free to use these sldes verbatm, or to modfy them to ft your

More information

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

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

More information

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

Some Advanced SPC Tools 1. Cumulative Sum Control (Cusum) Chart For the data shown in Table 9-1, the x chart can be generated.

Some Advanced SPC Tools 1. Cumulative Sum Control (Cusum) Chart For the data shown in Table 9-1, the x chart can be generated. Some Advanced SP Tools 1. umulatve Sum ontrol (usum) hart For the data shown n Table 9-1, the x chart can be generated. However, the shft taken place at sample #21 s not apparent. 92 For ths set samples,

More information

Querying by sketch geographical databases. Yu Han 1, a *

Querying by sketch geographical databases. Yu Han 1, a * 4th Internatonal Conference on Sensors, Measurement and Intellgent Materals (ICSMIM 2015) Queryng by sketch geographcal databases Yu Han 1, a * 1 Department of Basc Courses, Shenyang Insttute of Artllery,

More information

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning

Computer Animation and Visualisation. Lecture 4. Rigging / Skinning Computer Anmaton and Vsualsaton Lecture 4. Rggng / Sknnng Taku Komura Overvew Sknnng / Rggng Background knowledge Lnear Blendng How to decde weghts? Example-based Method Anatomcal models Sknnng Assume

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

Simulation: Solving Dynamic Models ABE 5646 Week 11 Chapter 2, Spring 2010

Simulation: Solving Dynamic Models ABE 5646 Week 11 Chapter 2, Spring 2010 Smulaton: Solvng Dynamc Models ABE 5646 Week Chapter 2, Sprng 200 Week Descrpton Readng Materal Mar 5- Mar 9 Evaluatng [Crop] Models Comparng a model wth data - Graphcal, errors - Measures of agreement

More information

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data

Type-2 Fuzzy Non-uniform Rational B-spline Model with Type-2 Fuzzy Data Malaysan Journal of Mathematcal Scences 11(S) Aprl : 35 46 (2017) Specal Issue: The 2nd Internatonal Conference and Workshop on Mathematcal Analyss (ICWOMA 2016) MALAYSIAN JOURNAL OF MATHEMATICAL SCIENCES

More information

Meta-heuristics for Multidimensional Knapsack Problems

Meta-heuristics for Multidimensional Knapsack Problems 2012 4th Internatonal Conference on Computer Research and Development IPCSIT vol.39 (2012) (2012) IACSIT Press, Sngapore Meta-heurstcs for Multdmensonal Knapsack Problems Zhbao Man + Computer Scence Department,

More information

Wishing you all a Total Quality New Year!

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

More information

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

Application of Learning Machine Methods to 3 D Object Modeling

Application of Learning Machine Methods to 3 D Object Modeling Applcaton of Learnng Machne Methods to 3 D Object Modelng Crstna Garca, and José Al Moreno Laboratoro de Computacón Emergente, Facultades de Cencas e Ingenería, Unversdad Central de Venezuela. Caracas,

More information

Problem Set 3 Solutions

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

More information

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS

A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Proceedngs of the Wnter Smulaton Conference M E Kuhl, N M Steger, F B Armstrong, and J A Jones, eds A MOVING MESH APPROACH FOR SIMULATION BUDGET ALLOCATION ON CONTINUOUS DOMAINS Mark W Brantley Chun-Hung

More information

APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT

APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT 3. - 5. 5., Brno, Czech Republc, EU APPLICATION OF MULTIVARIATE LOSS FUNCTION FOR ASSESSMENT OF THE QUALITY OF TECHNOLOGICAL PROCESS MANAGEMENT Abstract Josef TOŠENOVSKÝ ) Lenka MONSPORTOVÁ ) Flp TOŠENOVSKÝ

More information

Applying EM Algorithm for Segmentation of Textured Images

Applying EM Algorithm for Segmentation of Textured Images Proceedngs of the World Congress on Engneerng 2007 Vol I Applyng EM Algorthm for Segmentaton of Textured Images Dr. K Revathy, Dept. of Computer Scence, Unversty of Kerala, Inda Roshn V. S., ER&DCI Insttute

More information

Unsupervised Learning and Clustering

Unsupervised Learning and Clustering Unsupervsed Learnng and Clusterng Supervsed vs. Unsupervsed Learnng Up to now we consdered supervsed learnng scenaro, where we are gven 1. samples 1,, n 2. class labels for all samples 1,, n Ths s also

More information

BIN XIA et al: AN IMPROVED K-MEANS ALGORITHM BASED ON CLOUD PLATFORM FOR DATA MINING

BIN XIA et al: AN IMPROVED K-MEANS ALGORITHM BASED ON CLOUD PLATFORM FOR DATA MINING An Improved K-means Algorthm based on Cloud Platform for Data Mnng Bn Xa *, Yan Lu 2. School of nformaton and management scence, Henan Agrcultural Unversty, Zhengzhou, Henan 450002, P.R. Chna 2. College

More information

Correlative features for the classification of textural images

Correlative features for the classification of textural images Correlatve features for the classfcaton of textural mages M A Turkova 1 and A V Gadel 1, 1 Samara Natonal Research Unversty, Moskovskoe Shosse 34, Samara, Russa, 443086 Image Processng Systems Insttute

More information

Measuring Integration in the Network Structure: Some Suggestions on the Connectivity Index

Measuring Integration in the Network Structure: Some Suggestions on the Connectivity Index Measurng Integraton n the Network Structure: Some Suggestons on the Connectvty Inde 1. Measures of Connectvty The connectvty can be dvded nto two levels, one s domestc connectvty, n the case of the physcal

More information

Keyword-based Document Clustering

Keyword-based Document Clustering Keyword-based ocument lusterng Seung-Shk Kang School of omputer Scence Kookmn Unversty & AIrc hungnung-dong Songbuk-gu Seoul 36-72 Korea sskang@kookmn.ac.kr Abstract ocument clusterng s an aggregaton of

More information

Time Series Prediction Using RSOM and Local Models

Time Series Prediction Using RSOM and Local Models Tme Seres Predcton Usng RSOM and Local Models Peter ANGELOVIČ Slovak Unversty of Technology Faculty of Informatcs and Informaton Technologes Ilkovčova 3, 842 16 Bratslava, Slovaka angelovc@ft.stuba.sk

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

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

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

More information

Research on Categorization of Animation Effect Based on Data Mining

Research on Categorization of Animation Effect Based on Data Mining MATEC Web of Conferences 22, 0102 0 ( 2015) DOI: 10.1051/ matecconf/ 2015220102 0 C Owned by the authors, publshed by EDP Scences, 2015 Research on Categorzaton of Anmaton Effect Based on Data Mnng Na

More information

Machine Learning. Topic 6: Clustering

Machine Learning. Topic 6: Clustering Machne Learnng Topc 6: lusterng lusterng Groupng data nto (hopefully useful) sets. Thngs on the left Thngs on the rght Applcatons of lusterng Hypothess Generaton lusters mght suggest natural groups. Hypothess

More information

Solving two-person zero-sum game by Matlab

Solving two-person zero-sum game by Matlab Appled Mechancs and Materals Onlne: 2011-02-02 ISSN: 1662-7482, Vols. 50-51, pp 262-265 do:10.4028/www.scentfc.net/amm.50-51.262 2011 Trans Tech Publcatons, Swtzerland Solvng two-person zero-sum game by

More information

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

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

GSLM Operations Research II Fall 13/14

GSLM Operations Research II Fall 13/14 GSLM 58 Operatons Research II Fall /4 6. Separable Programmng Consder a general NLP mn f(x) s.t. g j (x) b j j =. m. Defnton 6.. The NLP s a separable program f ts objectve functon and all constrants are

More information

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

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

Selection of Streets from a Network Using Self-Organizing Maps

Selection of Streets from a Network Using Self-Organizing Maps NOT FOR CITATION, TO BE PUBLISHED IN TRANSACTIONS IN GIS 1 Selecton of Streets from a Network Usng Self-Organzng Maps Bn Jang 1 and Lars Harre 2 1 Dvson of Geomatcs, Dept. of Technology and Bult Envronment

More information

A fast algorithm for color image segmentation

A fast algorithm for color image segmentation Unersty of Wollongong Research Onlne Faculty of Informatcs - Papers (Arche) Faculty of Engneerng and Informaton Scences 006 A fast algorthm for color mage segmentaton L. Dong Unersty of Wollongong, lju@uow.edu.au

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

F Geometric Mean Graphs

F Geometric Mean Graphs Avalable at http://pvamu.edu/aam Appl. Appl. Math. ISSN: 1932-9466 Vol. 10, Issue 2 (December 2015), pp. 937-952 Applcatons and Appled Mathematcs: An Internatonal Journal (AAM) F Geometrc Mean Graphs A.

More information

Lecture 4: Principal components

Lecture 4: Principal components /3/6 Lecture 4: Prncpal components 3..6 Multvarate lnear regresson MLR s optmal for the estmaton data...but poor for handlng collnear data Covarance matrx s not nvertble (large condton number) Robustness

More information

CS 534: Computer Vision Model Fitting

CS 534: Computer Vision Model Fitting CS 534: Computer Vson Model Fttng Sprng 004 Ahmed Elgammal Dept of Computer Scence CS 534 Model Fttng - 1 Outlnes Model fttng s mportant Least-squares fttng Maxmum lkelhood estmaton MAP estmaton Robust

More information

Brave New World Pseudocode Reference

Brave New World Pseudocode Reference Brave New World Pseudocode Reference Pseudocode s a way to descrbe how to accomplsh tasks usng basc steps lke those a computer mght perform. In ths week s lab, you'll see how a form of pseudocode can be

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

Optimal Workload-based Weighted Wavelet Synopses

Optimal Workload-based Weighted Wavelet Synopses Optmal Workload-based Weghted Wavelet Synopses Yoss Matas School of Computer Scence Tel Avv Unversty Tel Avv 69978, Israel matas@tau.ac.l Danel Urel School of Computer Scence Tel Avv Unversty Tel Avv 69978,

More information

Biostatistics 615/815

Biostatistics 615/815 The E-M Algorthm Bostatstcs 615/815 Lecture 17 Last Lecture: The Smplex Method General method for optmzaton Makes few assumptons about functon Crawls towards mnmum Some recommendatons Multple startng ponts

More information

Optimal connection strategies in one- and two-dimensional associative memory models

Optimal connection strategies in one- and two-dimensional associative memory models Optmal connecton strateges n one- and two-dmensonal assocatve memory models Lee Calcraft, Rod Adams, and Nel Davey School of Computer Scence, Unversty of Hertfordshre College lane, Hatfeld, Hertfordshre

More information