A new algorithm to build feed forward neural networks.

Size: px
Start display at page:

Download "A new algorithm to build feed forward neural networks."

Transcription

1 A ew algorithm to build feed forward eural etworks. Amit Thombre Cetre of Excellece, Software Techologies ad Kowledge Maagemet, Tech Mahidra, Pue, Idia Abstract The paper presets a ew algorithm to build a feedforward eural etwork with a sigle hidde layer. The algorithm starts with 1 hidde uit ad the ew hidde uits are added to the etwork oly if they improve the classificatio accuracy of the etwork o the cross-validatio samples. The iitializatio of the weights ad bias is doe usig Nguye-Widrow method ad the etwork is traied without freezig the weights. The results show that the traiig accuracies obtaied usig this algorithm are better tha that obtaied usig N2C2S algorithm. The crossvalidatio accuracies ad the test predictio accuracies obtaied by usig both the algorithms are ot statistically sigificatly differet. Due to this ad also sice it is easy to uderstad ad implemet tha N2C2S algorithm, the proposed algorithm should be preferred tha the N2C2S algorithm. Alog with this 3 differet methods of obtaiig weights for eural etworks are also compared. The classificatio results obtaied usig this algorithm are compared to the predictio accuracies obtaied usig logistic regressio, C5.0 ad M5 classificatio techiques o 5 freely available data sets. The classificatio results show that NN is better tha logistic regressio over 2 data sets, equivalet i performace over 2 data sets ad has low performace tha logistic regressio i case of 1 data set. NN is better tha C5.0 over 1 data set ad equivalet i performace o the remaiig data sets. It is observed that M5 is a better classificatio techique tha other techiques over 1 dataset. M5 is better tha logistic regressio over 2 data sets ad equivalet i performace for the remaiig datasets. C5.0 is better tha logistic regressio over 1 data set ad equivalet i performace o the remaiig data sets. Keywords-classificatio, logistic regerssio, feedforward eural etwork, backpropagatio. I. INTRODUCTION Classificatio is a data miig (machie learig) techique which divides up data istaces such that each is assiged to oe of a umber of classes. The data istaces are assiged to precisely oe class ad ever to more tha oe class or ever to o class at all. Classificatio problems ca be foud i busiess, sciece, idustry, ad medicie. Some of the examples iclude bakruptcy predictio, customer chur predictio, credit scorig, medical diagosis (like predictig cacer), quality cotrol, hadwritte character recogitio, speech recogitio etc. Some of the widely used classificatio techiques are decisio trees, rule based classificatio, eural etworks (NN), Bayesia etworks, logistic regressio, k- earest eighbor classifier, support vector machies etc. I this study, the computatioally powerful techique based o Artificial Neural Net-works (ANN) is built usig a ew algorithm ad the the classificatio results are compared with those from logistic regressio, the model tree techique based o M5 ad the classical decisio tree based techique based o C5.0. NN have gaied popularity over a few years ad are beig successfully applied across wide problem domais such as fiace, medicie, egieerig, geology etc. NN ca adapt to the data without ay explicit specificatio of fuctioal or distributioal form for the uderlyig model [1]. NN ca approximate ay fuctio thus makig them flexible i modelig real world complex relatioships. Neural etworks have successfully bee applied for classificatio problems i bakruptcy predictio [2], [3], medical diagosis [4], [5], hadwritig recogitio [6] ad speech recogitio [7]. Artificial eural etworks are computig systems made up of large umber of simple, highly itercoected processig elemets called odes or artificial euros. The focus of this work is o the feedforward multilayer etworks or multilayer perceptros (MLPs) with 1 iput layer, 1 output layer ad 1 hidde layer as oe hidde layer is sufficiet to map a arbitrary fuctio to ay degree of accuracy. The umber of euros i the hidde layer eeds to be fixed to arrive at the correct architecture of the etwork. There are may algorithms which costruct the etworks with sigle hidde layer. I the Dyamic ode creatio (DNC) algorithm [8] the odes are added to the hidde layer oe at a time till a desired accuracy is obtaied. I Feed-forward Neural etwork Creatio Algorithm (FNNCA) [9] ad Costructive Algorithm for Real- Value Examples (CARVE)[10] the hidde uits are added to the hidde layer oe at a time util a etwork that completely recogizes all its iput patters is costructed. Usig these algorithms ca lead to overfittig of the traiig data ad do ot geeralize well with ukow data. I Neural etwork Costructio with Cross-Validatio Samples algorithm (N2C2S) [11] the odes are added to the hidde layer oly if they improve the accuracy of the etwork o the traiig ad the cross-validatio data. This algorithm uses freezig of weights which ca lead to icrease i umber of hidde uits [12]. I the proposed algorithm, the weights are ot froze ad the odes are added to the hidde layer oly if they improve the accuracy of the etwork o the cross-validatio data oly.

2 The cross-validatio data accuracy is selected because it is true measure of the performace of the model. Logistic regressio (sometimes called the logistic model or logit model) is used for predictio of the probability of occurrece of a evet by fittig data to a logit fuctio. The logistic fuctio which is as give below takes z as the iput ad f (z) is the output which is always betwee 0 ad 1. f (z)= e z /e z +1 =1/1+e z (1) The variable z comprises of other idepedet variables ad is give as z=β 0 +β 1 x 1 +β 2 x 2 +β 3 x 3 +β 4 x 4 +β 5 x β k x k... (2) where β0 is the itercept ad β1, β2, β3 are the regressio coefficiets of idepedet variables x1, x2, x3 respectively. Each of the regressio coefficiets describes the size of the cotributig idepedet variable. Logistic regressio is widely used i medical field [1], [13]. C5.0 is a algorithm to build decisio trees usig the cocept of iformatio etropy. The decisio tree is used as a predictive model which maps observatios about a item to coclusios about the item s target value. I these tree structures the leaves represet the class labels ad braches represet the rules of features that lead to those class labels. M5 builds tree based models i which the trees costructed have multivariate liear models. Thus these trees are aalogous to piecewise liear fuctios. They are applied to classificatio problems by applyig a stadard method of trasformig a classificatio problem ito a problem of fuctio approximatio [14]. M5 is based o M5 developed by Quila [15] but icludes techiques to hadle eumerated data ad missig values effectively [16]. The orgaizatio of the paper is as follows. After the itroductio i sectio I, sectio II presets the algorithm used for buildig the NN. Sectio III describes the experimetal setup. Sectio IV presets the results from the experimets ad sectio V gives the discussio ad coclusio from the work carried out. The paper cocludes with future work i Sectio VI. II. THE PROPOSED ALGORITHM The steps of the algorithm used to build the eural etwork are as follows:- 1. Let N 1 deote the etwork with I iput uits ad O output uits ad H hidde uits. 2. The iitializatio of the weights ad bias is doe usig Nguye-Widrow method [17]. This algorithm chooses values i order to distribute the active regio of each euro i the layer approximately evely across the layer s iput space. The weight values cotai a degree of radomess, so they are ot the same each time this fuctio is called. Also the wastage of euros by this method is less as compared to radom iitializatio [18]. 3. The accuracy of this etwork o the traiig ad validatio data set is A T1 ad A V1. 4. The umber of hidde uits is the icreased by 1 uit ad the weights are iitialized with the Nguye-Widrow method. I case of N2C2S the weights of the first H hidde uits of N 2 is obtaied from the optimal weights of N 1 ad the remaiig coectio weights are set radomly. Thus i the proposed algorithm the weights are ot froze ad the whole etwork is retraied as freezig requires large umber of hidde uits to achieve the same performace as that obtaied without freezig [12]. 5. Let the accuracy of this etwork N 2 be A T2 ad A V2 o the traiig ad the validatio set respectively. If A V2 > A V1 the N 2 etwork is better tha N 1 else the etwork is ru with icreasig the hidde uits by 1. Thus the steps 4 ad 5 are repeated till we get the etwork with highest accuracy over validatio data set. III. 1) Neural Networks EXPERIMENTAL SETUP a. The eural etwork was costructed usig Matlab. b. As metioed above i sectio I the etwork cosists of iput layer, output layer ad hidde layer. The umber of euros i the iput layer correspods to the umber of idepedet variables. The umber of euros i the hidde layer is obtaied by usig the algorithm metioed i sectio II. Ad the umber of euros i the output layer is correspodig to the category of classes i the depedet data mius oe for programmig purpose. c. The back propagatio traiig fuctio is scaled cojugate gradiet algorithm. d. Each ru of the etwork was carried out for 200 epochs with the goal for error set as 0. e. The Nguye-Widrow algorithm chooses the weight values i order to distribute the active regio of each euro i the layer approximately evely across the layer s iput space. These values cotai a degree of radomess, so they are ot the same each time this fuctio is called. That s why 30 rus were carried out for each fold of cross-validatio ad for each cofiguratio of the etwork to cosider as may as radom weight values ad the average of 30 rus is take. The 10 fold cross-validatio was ru 10 times ad the the predictio accuracy was averaged over these 10 rus. 2) Logistic Regressio The programmig for logistic regressio is doe usig R [19], a free statistical software. 10 times, 10 fold cross-validatio was used to carry out the rus. 3) C5.0 ad M5 The results usig these algorithms are obtaied from [14]. The experimets usig these algorithms were carried out usig 10 rus of te fold cross-validatio.

3 4) Data The Chapma data set was obtaied from [20] ad the rest of the data sets were obtaied from the UCI Machie Learig Laboratory [21]. The data sets were chose i such a way that they are publicly available ad the results of classificatio usig eural ets, C5.0 ad M5, o these data sets (except oe data set) is already available for compariso purpose. Also the data sets are small i size ad cotaied oly cotiuous data ad biary data. The missig cotiuous attribute value was replaced by the average of the o-missig values. The data was ormalized before carryig out the experimets o them. TABLE I. DETAILS OF THE DATA SETS Data sets Size Missig Attributes values Cotiuous Biary Nomial Glass(G2) Chapma Ioospher e Votig ) Experimets All the data sets used were split up ito 3 sets, viz traiig, validatio ad test data sets. Let them be referred as T R, V ad T E data sets respectively. Iitially the eural etwork is built usig the algorithm metioed i Sectio II. The umber of euros i the hidde layer varied from 1 to 20 maximum. These steps are commo to all the experimets metioed below. 1. Experimet A a. The weights i the commo step metioed above were obtaied correspodig to lowest geeralizatio error over validatio data set. b. The the etwork with weights fixed as obtaied from step a was ru o the traiig set comprisig of the traiig set T R ad validatio set V ad its accuracy was tested o the testig data set T E. This predictio accuracy is reported i the results. 2. Experimet B I this experimet, the steps a ad b remai the same as experimet A except that the criteria for obtaiig weights i step a is correspodig to the highest predictio accuracy over the validatio data set. 3. Experimet C I this experimet there was o exteral criteria laid dow as doe i experimets A ad B for gettig the weights. Thus the weights were ot fixed. The weights which were obtaied by each ru were used as it is for gettig the predictio accuracy over the traiig set. 1 * idicates that the data is ot available. IV. EXPERIMENTAL RESULTS The predictio accuracies obtaied by 2 differet methods were compared by usig the Welch's t test [22]. Welch s test is a adaptatio of Studet s t-test with the 2 samples havig uequal variaces. The ull hypothesis that the two meas are equal was rejected at the sigificace value of Followig are the results from various experimets tried out:- 1. The umber of hidde uits ad the accuracy rates of the eural etworks costructed by N2C2S ad the proposed algorithm are give i the tables II ad III below. Usig N2C2S Algorithm Data set Hidde uits Traiig Accuracy Crossvalidatio Accuracy Glass Chapma * 1 * * Ioosphe re Votig TABLE II. TABLE III. RESULTS USING N2C2S ALGORITHM Usig Metioed Algorithm Data set Hidde uits Traiig Accuracy Cross-validatio Accuracy Glass Chapma Ioosphe re Votig RESULTS USING MENTIONED ALGORITHM 2. The predictio accuracy over test data set obtaied usig experimets A, B ad C is give i the table below. This predictio accuracy is also compared with that obtaied from logistic regressio. TABLE IV PREDICTION ACCURACY BY NEURAL NETWORKS USING EXPERIMENTS A, B AND C Data Logistic Exp A Exp B Exp C sets Regressio Glass Chapma Ioosph ere Votig idicates that the predictio accuracy is less compared with other methods 3 idicates that the predictio accuracy is high compared with other methods

4 3. From table IV, the summary of results showig the compariso of eural etworks with logistic regressio is give i the table below. The wis ad losses are decided as per the Welch s test described at the start of this sectio. TABLE V. SUMMARY OF RESULTS NN versus Wi Ties Losses Logistic Regressio 4. The test predictio accuracies usig the N2C2S algorithm ad the proposed algorithm are give i the table below. TABLE VI. Data PREDICTION ACCURACIES USING N2C2S AND PROPOSED ALGORITHM Usig Exp A Exp B Exp C sets N2C2S Glass Chapma * Ioosphere Votig Referece [15] gives the predictio accuracies usig C5.0 ad M5 o the same data sets used i this study. The predictio accuracies of logistic regressio ad NN are compared with that obtaied by usig C5.0 ad M5 algorithms i the table below. The results of C5.0 ad M5 are the averages ad stadard deviatios from 10 rus of te fold cross-validatio experimets. The predictio accuracies usig NN from experimet A are used here. TABLE VII. PREDICTION ACCURACIES USING NN, LOGISTIC REGRESSION, C5.0 AND M5 Data sets Predictio Accuracy usig Neural Logistic C5.0 M5 Networks Regressio Glass Chapma * * Ioosphere Votig V. DISCUSSION AND CONCLUSIONS 1. Table VI shows that the predictio accuracies usig N2C2S is same as that obtaied usig the proposed algorithm. The proposed method of buildig the etwork used 30 iteratios for each ru, thus takig more combiatios of weights usig Nguye-Widrow iitializatio method ito cosideratio to arrive at the results as compared to ruig the N2C2S algorithm [11] which uses radom weight iitializatio oly oce. Also the results usig N2C2S algorithm are based o a smaller traiig set tha the oe used for the experimets carried out. Thus the results usig N2C2S algorithm are likely to chage if more iteratios ad a larger traiig set is cosidered. 2. The results i table II ad III show that the hidde uits obtaied usig proposed algorithm was less tha that obtaied usig N2C2S over 2 data sets ad greater i case of 2 other data sets. Thus it caot be cocluded if the ofreezig of weights has ay advatage over freezig of weights o the umber of hidde uits ad some more experimets eed to be performed over differet data sets to arrive at some coclusio. 3. The predictio accuracy over traiig data sets usig proposed algorithm was statistically better tha that obtaied usig N2C2S algorithm over 3 data sets. From tables II, III ad VI, it is observed that the cross-validatio accuracies ad test accuracies obtaied usig both the algorithms are ot sigificatly differet. Thus the proposed algorithm should be preferred as the chages are easy to uderstad ad implemet as compared to the N2C2S algorithm. 4. From table IV it is observed there is o sigle experimet which has performace better tha other 2 experimets. 5. From tables V it is observed that NN gives better performace tha logistic regressio o 2 data sets; gives the same results o 2 data sets ad lower tha that of logistic regressio o 1 data set. But it is observed that NN cosumes more executio time tha logistic regressio. Also logistic regressio techique is a white-box techique which allows the iterpretatio of the model parameters whereas NN is a black box techique which does ot allow the iterpretatio of the model. 6. NN is better tha C5.0 over 1 data set ad equivalet i performace o the remaiig data sets. 7. Table VII shows that M5 is better tha NN, logistic regressio ad C5.0 over 1 dataset. M5 is better tha logistic regressio over 2 data sets ad equivalet i performace for the remaiig datasets. M5 ad NN do ot have sigificatly differet accuracies except for 1 data set as metioed at the startig of this poit. M5 ad C5.0 do ot have sigificatly differet accuracies o the data sets except for 1as metioed at the startig of this poit. 8. From table VII, C5.0 is better tha logistic regressio over 1 data set ad equivalet i performace o the remaiig data sets. VI. FUTURE WORK From tables II, III ad VI, the proposed algorithm for buildig the eural etworks has give cross-validatio predictio accuracies ad test accuracies which are ot sigificatly differet with that obtaied by usig the N2C2S algorithm. The further work will be to improve this algorithm so that the

5 cross-validatio accuracies are better tha that obtaied usig N2C2S algorithm ad the predictio accuracies over test data are better tha that obtaied by usig N2C2S algorithm ad logistic regressio. Also some more experimets eed to be performed usig N2C2S algorithm ad the chages suggested i poit 1 of sectio V. ACKNOWLEDGEMENT I am grateful to Tech Mahidra Cetre of Excellece group for givig me the support to write this paper. REFERENCES [1] Guoqiag Peter Zhag, Neural Networks for classificatio: A Survey. IEEE Trasactios o Systems, Ma, ad Cyberetics- Part C: Applicatios ad reviews, vol. 30, No. 4, pp (2000) [2] Amir F. Atiya, Bakruptcy Predictio for Credit Risk Usig Neural Networks: A Survey ad New Results. IEEE Trasactios o Neural Networks, vol. 12, No. 4, pp (2000) [3] R. C. Lacher, P. K. Coats, S. C. Sharma, ad L. F. Fat, A eural etwork for classifyig the fiacial health of a firm. Eur. J. Oper. Res., vol. 85, pp (1995). [4] W. G. Baxt, Use of a artificial eural etwork for data aalysis i cliical decisio-makig: The diagosis of acute coroary occlusio. Neural Computig., vol. 2, pp (1990) [5] M. A. Mazurowski, P. A. Habas, J. M. Zurada, J.Y. Lo, J.A. Baker ad G. D. Tourassi, Traiig eural etwork classifiers for medical decisio makig: The effects of imbalaced datasets o classificatio performace. Neural Networks. vol. 21, pp (2007) [6] I. Guyo, Applicatios of eural etworks to character recogitio, Iteratioal Joural of Patter Recogitio ad Artificial. Itelligece. vol. 5, pp (1991) [7] H. Bourlard ad N. Morga, Cotiuous speech recogitio by coectioist statistical methods. IEEE Trasactios o Neural Networks. vol. 4, pp (1993) [8] Ash, T, Dyamic ode creatio i backpropagatio etworks. Coectio Sciece. 1(4), (2002) [9] Rudy Setioo, Feedforward Neural Network Costructio Usig Cross Validatio. Neural Computatio. 13:12, (2001) [10] Youg, S. ad Dows, T, CARVE -a costructive algorithm for realvalued examples. IEEE Trasactio o Neural Networks 9(6), (1998) [11] Setioo R.: A Neural Network Costructio Algorithm which Maximizes the Likelihood Fuctio. Coectio Sciece. vol 7, Number 2, pp (1996) [12] Ti-Yau Kwok ad Dit-Ya Yeug, Experimetal aalysis of iput weight freezig i costructig eural etworks, IEEE Iteratioal Coferece o Neural Networks. vol 1, pp (1993) [13] J.G. Liao ad Chi, Logistic regressio for disease classificatio usig microarray data: model selectio i a large p ad small case. Bioiformatics, vol 23, pp (2007) [14] Frak, E., Wag, Y, Iglis, S., Holmes, G. ad Witte, I.H, Usig Model trees for classificatio. Machie Learig 32(1), pp (1997) [15] Quila, J.R. Learig with cotiuous classes, Proceedigs Australia Joit Coferece o Artificial Itelligece (pp ). World Scietific, Sigapore(1992). [16] Wag, Y. & Witte, I.H. (1997). Iductio of model trees for predictig cotiuous classes, Proceedigs of the poster papers of the Europea Coferece o Machie Learig, Uiversity of Ecoomics, Faculty of Iformatics ad Statistics, Prague. [17] Derrick Nguye ad Berard Widrow, Improvig the learig speed of 2-layer eural etworks by choosig iitial values of the adaptive weights. Proceedigs of the Iteratioal Joit Coferece o Neural Networks, 3: (1990) [18] Howard Demuth, Mark Beale ad Marti T. Haga, Neural Network Toolbox 7, User s Guide, The MathWorks, Ic., Natick, MA, Revised for Versio 7.0 (Release 2010b), September [19] [20] [21] UCI Machie learig repository data sets, [22] Welch, B. L, The geeralizatio of "Studet's" problem whe several differet populatio variaces are ivolved. Biometrika 34 (1 2): (1947)

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON

A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON A SOFTWARE MODEL FOR THE MULTILAYER PERCEPTRON Roberto Lopez ad Eugeio Oñate Iteratioal Ceter for Numerical Methods i Egieerig (CIMNE) Edificio C1, Gra Capitá s/, 08034 Barceloa, Spai ABSTRACT I this work

More information

Investigating methods for improving Bagged k-nn classifiers

Investigating methods for improving Bagged k-nn classifiers Ivestigatig methods for improvig Bagged k-nn classifiers Fuad M. Alkoot Telecommuicatio & Navigatio Istitute, P.A.A.E.T. P.O.Box 4575, Alsalmia, 22046 Kuwait Abstract- We experimet with baggig knn classifiers

More information

ANN WHICH COVERS MLP AND RBF

ANN WHICH COVERS MLP AND RBF ANN WHICH COVERS MLP AND RBF Josef Boští, Jaromír Kual Faculty of Nuclear Scieces ad Physical Egieerig, CTU i Prague Departmet of Software Egieerig Abstract Two basic types of artificial eural etwors Multi

More information

Designing a learning system

Designing a learning system CS 75 Itro to Machie Learig Lecture Desigig a learig system Milos Hauskrecht milos@pitt.edu 539 Seott Square, -5 people.cs.pitt.edu/~milos/courses/cs75/ Admiistrivia No homework assigmet this week Please

More information

Pattern Recognition Systems Lab 1 Least Mean Squares

Pattern Recognition Systems Lab 1 Least Mean Squares Patter Recogitio Systems Lab 1 Least Mea Squares 1. Objectives This laboratory work itroduces the OpeCV-based framework used throughout the course. I this assigmet a lie is fitted to a set of poits usig

More information

Designing a learning system

Designing a learning system CS 75 Machie Learig Lecture Desigig a learig system Milos Hauskrecht milos@cs.pitt.edu 539 Seott Square, x-5 people.cs.pitt.edu/~milos/courses/cs75/ Admiistrivia No homework assigmet this week Please try

More information

3D Model Retrieval Method Based on Sample Prediction

3D Model Retrieval Method Based on Sample Prediction 20 Iteratioal Coferece o Computer Commuicatio ad Maagemet Proc.of CSIT vol.5 (20) (20) IACSIT Press, Sigapore 3D Model Retrieval Method Based o Sample Predictio Qigche Zhag, Ya Tag* School of Computer

More information

Lecture 13: Validation

Lecture 13: Validation Lecture 3: Validatio Resampli methods Holdout Cross Validatio Radom Subsampli -Fold Cross-Validatio Leave-oe-out The Bootstrap Bias ad variace estimatio Three-way data partitioi Itroductio to Patter Recoitio

More information

Euclidean Distance Based Feature Selection for Fault Detection Prediction Model in Semiconductor Manufacturing Process

Euclidean Distance Based Feature Selection for Fault Detection Prediction Model in Semiconductor Manufacturing Process Vol.133 (Iformatio Techology ad Computer Sciece 016), pp.85-89 http://dx.doi.org/10.1457/astl.016. Euclidea Distace Based Feature Selectio for Fault Detectio Predictio Model i Semicoductor Maufacturig

More information

Our second algorithm. Comp 135 Machine Learning Computer Science Tufts University. Decision Trees. Decision Trees. Decision Trees.

Our second algorithm. Comp 135 Machine Learning Computer Science Tufts University. Decision Trees. Decision Trees. Decision Trees. Comp 135 Machie Learig Computer Sciece Tufts Uiversity Fall 2017 Roi Khardo Some of these slides were adapted from previous slides by Carla Brodley Our secod algorithm Let s look at a simple dataset for

More information

EMPIRICAL ANALYSIS OF FAULT PREDICATION TECHNIQUES FOR IMPROVING SOFTWARE PROCESS CONTROL

EMPIRICAL ANALYSIS OF FAULT PREDICATION TECHNIQUES FOR IMPROVING SOFTWARE PROCESS CONTROL Iteratioal Joural of Iformatio Techology ad Kowledge Maagemet July-December 2012, Volume 5, No. 2, pp. 371-375 EMPIRICAL ANALYSIS OF FAULT PREDICATION TECHNIQUES FOR IMPROVING SOFTWARE PROCESS CONTROL

More information

Dynamic Programming and Curve Fitting Based Road Boundary Detection

Dynamic Programming and Curve Fitting Based Road Boundary Detection Dyamic Programmig ad Curve Fittig Based Road Boudary Detectio SHYAM PRASAD ADHIKARI, HYONGSUK KIM, Divisio of Electroics ad Iformatio Egieerig Chobuk Natioal Uiversity 664-4 Ga Deokji-Dog Jeoju-City Jeobuk

More information

Transactions of the VŠB Technical University of Ostrava, Mechanical Series. article No. 1965

Transactions of the VŠB Technical University of Ostrava, Mechanical Series. article No. 1965 Trasactios of the VŠB Techical Uiversity of Ostrava, Mechaical Series No. 2, 2013, vol. LIX article No. 1965 Marti MARIŠKA *, Petr DOLEŽEL ** PIECEWISE-LINEAR NEURAL NETWORK POSSIBLE TRAINING ALGORITHMS

More information

Probabilistic Fuzzy Time Series Method Based on Artificial Neural Network

Probabilistic Fuzzy Time Series Method Based on Artificial Neural Network America Joural of Itelliget Systems 206, 6(2): 42-47 DOI: 0.5923/j.ajis.2060602.02 Probabilistic Fuzzy Time Series Method Based o Artificial Neural Network Erol Egrioglu,*, Ere Bas, Cagdas Haka Aladag

More information

New HSL Distance Based Colour Clustering Algorithm

New HSL Distance Based Colour Clustering Algorithm The 4th Midwest Artificial Itelligece ad Cogitive Scieces Coferece (MAICS 03 pp 85-9 New Albay Idiaa USA April 3-4 03 New HSL Distace Based Colour Clusterig Algorithm Vasile Patrascu Departemet of Iformatics

More information

Bayesian approach to reliability modelling for a probability of failure on demand parameter

Bayesian approach to reliability modelling for a probability of failure on demand parameter Bayesia approach to reliability modellig for a probability of failure o demad parameter BÖRCSÖK J., SCHAEFER S. Departmet of Computer Architecture ad System Programmig Uiversity Kassel, Wilhelmshöher Allee

More information

DATA MINING II - 1DL460

DATA MINING II - 1DL460 DATA MINING II - 1DL460 Sprig 2017 A secod course i data miig http://www.it.uu.se/edu/course/homepage/ifoutv2/vt17/ Kjell Orsbor Uppsala Database Laboratory Departmet of Iformatio Techology, Uppsala Uiversity,

More information

x x 2 x Iput layer = quatity of classificatio mode X T = traspositio matrix The core of such coditioal probability estimatig method is calculatig the

x x 2 x Iput layer = quatity of classificatio mode X T = traspositio matrix The core of such coditioal probability estimatig method is calculatig the COMPARATIVE RESEARCHES ON PROBABILISTIC NEURAL NETWORKS AND MULTI-LAYER PERCEPTRON NETWORKS FOR REMOTE SENSING IMAGE SEGMENTATION Liu Gag a, b, * a School of Electroic Iformatio, Wuha Uiversity, 430079,

More information

New Fuzzy Color Clustering Algorithm Based on hsl Similarity

New Fuzzy Color Clustering Algorithm Based on hsl Similarity IFSA-EUSFLAT 009 New Fuzzy Color Clusterig Algorithm Based o hsl Similarity Vasile Ptracu Departmet of Iformatics Techology Tarom Compay Bucharest Romaia Email: patrascu.v@gmail.com Abstract I this paper

More information

Intrusion Detection using Fuzzy Clustering and Artificial Neural Network

Intrusion Detection using Fuzzy Clustering and Artificial Neural Network Itrusio Detectio usig Fuzzy Clusterig ad Artificial Neural Network Shraddha Suraa Research Scholar, Departmet of Computer Egieerig, Vishwakarma Istitute of Techology, Pue Idia shraddha.suraa@gmail.com

More information

(3) The denominator is ineffective for making decisions. Because C is the same for all values.on the other hand due to our independent feature set

(3) The denominator is ineffective for making decisions. Because C is the same for all values.on the other hand due to our independent feature set ORIGINAL ARTICLE Received 17 Ja. 2014 Accepted 30 Ja. 2014 2014, Sciecelie Publicatio www.sciece-lie.com 2322-5114 Joural of World s Electrical Egieerig ad Techology J. World. Elect. Eg. Tech. 3(1): 01-05,

More information

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem

An Improved Shuffled Frog-Leaping Algorithm for Knapsack Problem A Improved Shuffled Frog-Leapig Algorithm for Kapsack Problem Zhoufag Li, Ya Zhou, ad Peg Cheg School of Iformatio Sciece ad Egieerig Hea Uiversity of Techology ZhegZhou, Chia lzhf1978@126.com Abstract.

More information

Cubic Polynomial Curves with a Shape Parameter

Cubic Polynomial Curves with a Shape Parameter roceedigs of the th WSEAS Iteratioal Coferece o Robotics Cotrol ad Maufacturig Techology Hagzhou Chia April -8 00 (pp5-70) Cubic olyomial Curves with a Shape arameter MO GUOLIANG ZHAO YANAN Iformatio ad

More information

CLASSIFICATION MODELS BASED FORWARD SELECTION FOR BUSINESS PERFORMANCE PREDICTION

CLASSIFICATION MODELS BASED FORWARD SELECTION FOR BUSINESS PERFORMANCE PREDICTION Joural of Theoretical ad Applied Iformatio Techology 20 th September 204. Vol. 67 No.2 2005-204 JATIT & LLS. All rights reserved. ISSN: 992-8645 www.atit.org E-ISSN: 87-395 CLASSIFICATION MODELS BASED

More information

Parallel Polygon Approximation Algorithm Targeted at Reconfigurable Multi-Ring Hardware

Parallel Polygon Approximation Algorithm Targeted at Reconfigurable Multi-Ring Hardware Parallel Polygo Approximatio Algorithm Targeted at Recofigurable Multi-Rig Hardware M. Arif Wai* ad Hamid R. Arabia** *Califoria State Uiversity Bakersfield, Califoria, USA **Uiversity of Georgia, Georgia,

More information

. Written in factored form it is easy to see that the roots are 2, 2, i,

. Written in factored form it is easy to see that the roots are 2, 2, i, CMPS A Itroductio to Programmig Programmig Assigmet 4 I this assigmet you will write a java program that determies the real roots of a polyomial that lie withi a specified rage. Recall that the roots (or

More information

Improving Template Based Spike Detection

Improving Template Based Spike Detection Improvig Template Based Spike Detectio Kirk Smith, Member - IEEE Portlad State Uiversity petra@ee.pdx.edu Abstract Template matchig algorithms like SSE, Covolutio ad Maximum Likelihood are well kow for

More information

Neural Networks A Model of Boolean Functions

Neural Networks A Model of Boolean Functions Neural Networks A Model of Boolea Fuctios Berd Steibach, Roma Kohut Freiberg Uiversity of Miig ad Techology Istitute of Computer Sciece D-09596 Freiberg, Germay e-mails: steib@iformatik.tu-freiberg.de

More information

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve

Analysis of Server Resource Consumption of Meteorological Satellite Application System Based on Contour Curve Advaces i Computer, Sigals ad Systems (2018) 2: 19-25 Clausius Scietific Press, Caada Aalysis of Server Resource Cosumptio of Meteorological Satellite Applicatio System Based o Cotour Curve Xiagag Zhao

More information

Our Learning Problem, Again

Our Learning Problem, Again Noparametric Desity Estimatio Matthew Stoe CS 520, Sprig 2000 Lecture 6 Our Learig Problem, Agai Use traiig data to estimate ukow probabilities ad probability desity fuctios So far, we have depeded o describig

More information

A Novel Hybrid Algorithm for Software Cost Estimation Based on Cuckoo Optimization and K-Nearest Neighbors Algorithms

A Novel Hybrid Algorithm for Software Cost Estimation Based on Cuckoo Optimization and K-Nearest Neighbors Algorithms Egieerig, Techology & Applied Sciece Research Vol. 6, No. 3, 26, 8-22 8 A Novel Hybrid Algorithm for Software Cost Estimatio Based o Optimizatio ad K-Nearest Neighbors Algorithms Elaz Eskadaria Miadoab

More information

Speeding-up dynamic programming in sequence alignment

Speeding-up dynamic programming in sequence alignment Departmet of Computer Sciece Aarhus Uiversity Demark Speedig-up dyamic programmig i sequece aligmet Master s Thesis Dug My Hoa - 443 December, Supervisor: Christia Nørgaard Storm Pederse Implemetatio code

More information

Criterion in selecting the clustering algorithm in Radial Basis Functional Link Nets

Criterion in selecting the clustering algorithm in Radial Basis Functional Link Nets WSEAS TRANSACTIONS o SYSTEMS Ag Sau Loog, Og Hog Choo, Low Heg Chi Criterio i selectig the clusterig algorithm i Radial Basis Fuctioal Lik Nets ANG SAU LOONG 1, ONG HONG CHOON 2 & LOW HENG CHIN 3 Departmet

More information

Comparison of classification algorithms in the task of object recognition on radar images of the MSTAR base

Comparison of classification algorithms in the task of object recognition on radar images of the MSTAR base Compariso of classificatio algorithms i the task of object recogitio o radar images of the MSTAR base A.A. Borodiov 1, V.V. Myasikov 1,2 1 Samara Natioal Research Uiversity, 34 Moskovskoe Shosse, 443086,

More information

Performance Evaluation of Mutation / Non- Mutation Based Classification With Missing Data

Performance Evaluation of Mutation / Non- Mutation Based Classification With Missing Data Performace Evaluatio of Mutatio / No- Mutatio Based Classificatio With Missig Data N.C. Viod Research Scholar, Maomaiam Sudaraar Uiversity, Tiruelveli, Tamil Nadu, Idia Dr. M. Puithavalli Research Supervisor,

More information

EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS

EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS Iteratioal Joural o Natural Laguage Computig (IJNLC) Vol. 2, No., February 203 EFFECT OF QUERY FORMATION ON WEB SEARCH ENGINE RESULTS Raj Kishor Bisht ad Ila Pat Bisht 2 Departmet of Computer Sciece &

More information

Fundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le

Fundamentals of Media Processing. Shin'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dinh Le Fudametals of Media Processig Shi'ichi Satoh Kazuya Kodama Hiroshi Mo Duy-Dih Le Today's topics Noparametric Methods Parze Widow k-nearest Neighbor Estimatio Clusterig Techiques k-meas Agglomerative Hierarchical

More information

Evaluation of Support Vector Machine Kernels for Detecting Network Anomalies

Evaluation of Support Vector Machine Kernels for Detecting Network Anomalies Evaluatio of Support Vector Machie Kerels for Detectig Network Aomalies Prera Batta, Maider Sigh, Zhida Li, Qigye Dig, ad Ljiljaa Trajković Commuicatio Networks Laboratory http://www.esc.sfu.ca/~ljilja/cl/

More information

Analysis of Documents Clustering Using Sampled Agglomerative Technique

Analysis of Documents Clustering Using Sampled Agglomerative Technique Aalysis of Documets Clusterig Usig Sampled Agglomerative Techique Omar H. Karam, Ahmed M. Hamad, ad Sheri M. Moussa Abstract I this paper a clusterig algorithm for documets is proposed that adapts a samplig-based

More information

Handwriting Stroke Extraction Using a New XYTC Transform

Handwriting Stroke Extraction Using a New XYTC Transform Hadwritig Stroke Etractio Usig a New XYTC Trasform Gilles F. Houle 1, Kateria Bliova 1 ad M. Shridhar 1 Computer Scieces Corporatio Uiversity Michiga-Dearbor Abstract: The fudametal represetatio of hadwritig

More information

Chapter 1. Introduction to Computers and C++ Programming. Copyright 2015 Pearson Education, Ltd.. All rights reserved.

Chapter 1. Introduction to Computers and C++ Programming. Copyright 2015 Pearson Education, Ltd.. All rights reserved. Chapter 1 Itroductio to Computers ad C++ Programmig Copyright 2015 Pearso Educatio, Ltd.. All rights reserved. Overview 1.1 Computer Systems 1.2 Programmig ad Problem Solvig 1.3 Itroductio to C++ 1.4 Testig

More information

Study on effective detection method for specific data of large database LI Jin-feng

Study on effective detection method for specific data of large database LI Jin-feng Iteratioal Coferece o Automatio, Mechaical Cotrol ad Computatioal Egieerig (AMCCE 205) Study o effective detectio method for specific data of large database LI Ji-feg (Vocatioal College of DogYig, Shadog

More information

Fuzzy Rule Selection by Data Mining Criteria and Genetic Algorithms

Fuzzy Rule Selection by Data Mining Criteria and Genetic Algorithms Fuzzy Rule Selectio by Data Miig Criteria ad Geetic Algorithms Hisao Ishibuchi Dept. of Idustrial Egieerig Osaka Prefecture Uiversity 1-1 Gakue-cho, Sakai, Osaka 599-8531, JAPAN E-mail: hisaoi@ie.osakafu-u.ac.jp

More information

Mining from Quantitative Data with Linguistic Minimum Supports and Confidences

Mining from Quantitative Data with Linguistic Minimum Supports and Confidences Miig from Quatitative Data with Liguistic Miimum Supports ad Cofideces Tzug-Pei Hog, Mig-Jer Chiag ad Shyue-Liag Wag Departmet of Electrical Egieerig Natioal Uiversity of Kaohsiug Kaohsiug, 8, Taiwa, R.O.C.

More information

A Parallel DFA Minimization Algorithm

A Parallel DFA Minimization Algorithm A Parallel DFA Miimizatio Algorithm Ambuj Tewari, Utkarsh Srivastava, ad P. Gupta Departmet of Computer Sciece & Egieerig Idia Istitute of Techology Kapur Kapur 208 016,INDIA pg@iitk.ac.i Abstract. I this

More information

Neuro Fuzzy Model for Human Face Expression Recognition

Neuro Fuzzy Model for Human Face Expression Recognition IOSR Joural of Computer Egieerig (IOSRJCE) ISSN : 2278-0661 Volume 1, Issue 2 (May-Jue 2012), PP 01-06 Neuro Fuzzy Model for Huma Face Expressio Recogitio Mr. Mayur S. Burage 1, Prof. S. V. Dhopte 2 1

More information

arxiv: v2 [cs.ds] 24 Mar 2018

arxiv: v2 [cs.ds] 24 Mar 2018 Similar Elemets ad Metric Labelig o Complete Graphs arxiv:1803.08037v [cs.ds] 4 Mar 018 Pedro F. Felzeszwalb Brow Uiversity Providece, RI, USA pff@brow.edu March 8, 018 We cosider a problem that ivolves

More information

Revisiting the performance of mixtures of software reliability growth models

Revisiting the performance of mixtures of software reliability growth models Revisitig the performace of mixtures of software reliability growth models Peter A. Keiller 1, Charles J. Kim 1, Joh Trimble 1, ad Marlo Mejias 2 1 Departmet of Systems ad Computer Sciece 2 Departmet of

More information

Reversible Realization of Quaternary Decoder, Multiplexer, and Demultiplexer Circuits

Reversible Realization of Quaternary Decoder, Multiplexer, and Demultiplexer Circuits Egieerig Letters, :, EL Reversible Realizatio of Quaterary Decoder, Multiplexer, ad Demultiplexer Circuits Mozammel H.. Kha, Member, ENG bstract quaterary reversible circuit is more compact tha the correspodig

More information

Clustering and Classifying Diabetic Data Sets Using K-Means Algorithm

Clustering and Classifying Diabetic Data Sets Using K-Means Algorithm Article ca be accessed olie at http://www.publishigidia.com Clusterig ad Classifyig Diabetic Data Sets Usig K-Meas Algorithm M. Kothaiayaki*, P. Thagaraj** Abstract The k-meas algorithm is well kow for

More information

IMP: Superposer Integrated Morphometrics Package Superposition Tool

IMP: Superposer Integrated Morphometrics Package Superposition Tool IMP: Superposer Itegrated Morphometrics Package Superpositio Tool Programmig by: David Lieber ( 03) Caisius College 200 Mai St. Buffalo, NY 4208 Cocept by: H. David Sheets, Dept. of Physics, Caisius College

More information

Redundancy Allocation for Series Parallel Systems with Multiple Constraints and Sensitivity Analysis

Redundancy Allocation for Series Parallel Systems with Multiple Constraints and Sensitivity Analysis IOSR Joural of Egieerig Redudacy Allocatio for Series Parallel Systems with Multiple Costraits ad Sesitivity Aalysis S. V. Suresh Babu, D.Maheswar 2, G. Ragaath 3 Y.Viaya Kumar d G.Sakaraiah e (Mechaical

More information

An Anomaly Detection Method Based On Deep Learning

An Anomaly Detection Method Based On Deep Learning Sed Orders for Reprits to reprits@bethamsciece.ae 734 The Ope Automatio ad Cotrol Systems Joural, 05, 7, 734-739 A Aomaly Detectio Method Based O Deep Learig Ope Access Hog-li Deg *, Tao yag ad Jiag-i

More information

A Novel Feature Extraction Algorithm for Haar Local Binary Pattern Texture Based on Human Vision System

A Novel Feature Extraction Algorithm for Haar Local Binary Pattern Texture Based on Human Vision System A Novel Feature Extractio Algorithm for Haar Local Biary Patter Texture Based o Huma Visio System Liu Tao 1,* 1 Departmet of Electroic Egieerig Shaaxi Eergy Istitute Xiayag, Shaaxi, Chia Abstract The locality

More information

HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING

HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING Y.K. Patil* Iteratioal Joural of Advaced Research i ISSN: 2278-6244 IT ad Egieerig Impact Factor: 4.54 HADOOP: A NEW APPROACH FOR DOCUMENT CLUSTERING Prof. V.S. Nadedkar** Abstract: Documet clusterig is

More information

Decision Support Systems

Decision Support Systems Decisio Support Systems 50 (010) 93 10 Cotets lists available at ScieceDirect Decisio Support Systems joural homepage: www.elsevier.com/locate/dss The data complexity idex to costruct a efficiet cross-validatio

More information

Pruning and Summarizing the Discovered Time Series Association Rules from Mechanical Sensor Data Qing YANG1,a,*, Shao-Yu WANG1,b, Ting-Ting ZHANG2,c

Pruning and Summarizing the Discovered Time Series Association Rules from Mechanical Sensor Data Qing YANG1,a,*, Shao-Yu WANG1,b, Ting-Ting ZHANG2,c Advaces i Egieerig Research (AER), volume 131 3rd Aual Iteratioal Coferece o Electroics, Electrical Egieerig ad Iformatio Sciece (EEEIS 2017) Pruig ad Summarizig the Discovered Time Series Associatio Rules

More information

1 Graph Sparsfication

1 Graph Sparsfication CME 305: Discrete Mathematics ad Algorithms 1 Graph Sparsficatio I this sectio we discuss the approximatio of a graph G(V, E) by a sparse graph H(V, F ) o the same vertex set. I particular, we cosider

More information

BASED ON ITERATIVE ERROR-CORRECTION

BASED ON ITERATIVE ERROR-CORRECTION A COHPARISO OF CRYPTAALYTIC PRICIPLES BASED O ITERATIVE ERROR-CORRECTIO Miodrag J. MihaljeviC ad Jova Dj. GoliC Istitute of Applied Mathematics ad Electroics. Belgrade School of Electrical Egieerig. Uiversity

More information

Classification by Evolutionary Generalized Radial Basis Functions

Classification by Evolutionary Generalized Radial Basis Functions 2009 Nith teratioal Coferece o telliget Systems Desig ad Applicatios Classificatio by Evolutioary Geeralized Radial Basis Fuctios A. Castaño, C. Hervás-Martíez 2, P. A. Gutierrez 2, F. Ferádez-Navarro

More information

Enhancing Efficiency of Software Fault Tolerance Techniques in Satellite Motion System

Enhancing Efficiency of Software Fault Tolerance Techniques in Satellite Motion System Joural of Iformatio Systems ad Telecommuicatio, Vol. 2, No. 3, July-September 2014 173 Ehacig Efficiecy of Software Fault Tolerace Techiques i Satellite Motio System Hoda Baki Departmet of Electrical ad

More information

Using The Central Limit Theorem for Belief Network Learning

Using The Central Limit Theorem for Belief Network Learning Usig The Cetral Limit Theorem for Belief Network Learig Ia Davidso, Mioo Amiia Computer Sciece Dept, SUNY Albay Albay, NY, USA,. davidso@cs.albay.edu Abstract. Learig the parameters (coditioal ad margial

More information

Data diverse software fault tolerance techniques

Data diverse software fault tolerance techniques Data diverse software fault tolerace techiques Complemets desig diversity by compesatig for desig diversity s s limitatios Ivolves obtaiig a related set of poits i the program data space, executig the

More information

Sorting in Linear Time. Data Structures and Algorithms Andrei Bulatov

Sorting in Linear Time. Data Structures and Algorithms Andrei Bulatov Sortig i Liear Time Data Structures ad Algorithms Adrei Bulatov Algorithms Sortig i Liear Time 7-2 Compariso Sorts The oly test that all the algorithms we have cosidered so far is compariso The oly iformatio

More information

Software Fault Prediction of Unlabeled Program Modules

Software Fault Prediction of Unlabeled Program Modules Software Fault Predictio of Ulabeled Program Modules C. Catal, U. Sevim, ad B. Diri, Member, IAENG Abstract Software metrics ad fault data belogig to a previous software versio are used to build the software

More information

Task scenarios Outline. Scenarios in Knowledge Extraction. Proposed Framework for Scenario to Design Diagram Transformation

Task scenarios Outline. Scenarios in Knowledge Extraction. Proposed Framework for Scenario to Design Diagram Transformation 6-0-0 Kowledge Trasformatio from Task Scearios to View-based Desig Diagrams Nima Dezhkam Kamra Sartipi {dezhka, sartipi}@mcmaster.ca Departmet of Computig ad Software McMaster Uiversity CANADA SEKE 08

More information

An Efficient Algorithm for Graph Bisection of Triangularizations

An Efficient Algorithm for Graph Bisection of Triangularizations A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045 Oe Brookigs Drive St. Louis, Missouri 63130-4899, USA jaegerg@cse.wustl.edu

More information

EM375 STATISTICS AND MEASUREMENT UNCERTAINTY LEAST SQUARES LINEAR REGRESSION ANALYSIS

EM375 STATISTICS AND MEASUREMENT UNCERTAINTY LEAST SQUARES LINEAR REGRESSION ANALYSIS EM375 STATISTICS AND MEASUREMENT UNCERTAINTY LEAST SQUARES LINEAR REGRESSION ANALYSIS I this uit of the course we ivestigate fittig a straight lie to measured (x, y) data pairs. The equatio we wat to fit

More information

A Two-Level Approach to Making Class Predictions

A Two-Level Approach to Making Class Predictions A Two-Level Approach to Makig Class Predictios Adria Costea Turku Cetre for Computer Sciece ad IAMSR / Åbo Akademi Uiversity, Turku, Filad, Adria.Costea@abo.fi Tomas Eklud Turku Cetre for Computer Sciece

More information

Empirical Validate C&K Suite for Predict Fault-Proneness of Object-Oriented Classes Developed Using Fuzzy Logic.

Empirical Validate C&K Suite for Predict Fault-Proneness of Object-Oriented Classes Developed Using Fuzzy Logic. Empirical Validate C&K Suite for Predict Fault-Proeess of Object-Orieted Classes Developed Usig Fuzzy Logic. Mohammad Amro 1, Moataz Ahmed 1, Kaaa Faisal 2 1 Iformatio ad Computer Sciece Departmet, Kig

More information

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS

FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS FREQUENCY ESTIMATION OF INTERNET PACKET STREAMS WITH LIMITED SPACE: UPPER AND LOWER BOUNDS Prosejit Bose Evagelos Kraakis Pat Mori Yihui Tag School of Computer Sciece, Carleto Uiversity {jit,kraakis,mori,y

More information

Pseudocode ( 1.1) Analysis of Algorithms. Primitive Operations. Pseudocode Details. Running Time ( 1.1) Estimating performance

Pseudocode ( 1.1) Analysis of Algorithms. Primitive Operations. Pseudocode Details. Running Time ( 1.1) Estimating performance Aalysis of Algorithms Iput Algorithm Output A algorithm is a step-by-step procedure for solvig a problem i a fiite amout of time. Pseudocode ( 1.1) High-level descriptio of a algorithm More structured

More information

Effect of control points distribution on the orthorectification accuracy of an Ikonos II image through rational polynomial functions

Effect of control points distribution on the orthorectification accuracy of an Ikonos II image through rational polynomial functions Effect of cotrol poits distributio o the orthorectificatio accuracy of a Ikoos II image through ratioal polyomial fuctios Marcela do Valle Machado 1, Mauro Homem Atues 1 ad Paula Debiasi 1 1 Federal Rural

More information

A Comparative Study on Modeling Methods for Switched Reluctance Machines

A Comparative Study on Modeling Methods for Switched Reluctance Machines Computer ad Iformatio Sciece Vol. 3, No. 2; May 21 A Comparative Study o Modelig Methods for Switched Reluctace Machies Shouju Sog & Weiguo iu School of Automatio, Northwester Polytechical Uiversity PO

More information

A Wrapper-Based Combined Recursive Orthogonal Array and Support Vector Machine for Classification and Feature Selection

A Wrapper-Based Combined Recursive Orthogonal Array and Support Vector Machine for Classification and Feature Selection Moder Applied Sciece; Vol. 8, No. ; 24 ISSN 93-844 E-ISSN 93-852 Published by Caadia Ceter of Sciece ad Educatio A Wrapper-Based Combied Recursive Orthogoal Array ad Support Vector Machie for Classificatio

More information

Improving Information Retrieval System Security via an Optimal Maximal Coding Scheme

Improving Information Retrieval System Security via an Optimal Maximal Coding Scheme Improvig Iformatio Retrieval System Security via a Optimal Maximal Codig Scheme Dogyag Log Departmet of Computer Sciece, City Uiversity of Hog Kog, 8 Tat Chee Aveue Kowloo, Hog Kog SAR, PRC dylog@cs.cityu.edu.hk

More information

CIS 121 Data Structures and Algorithms with Java Spring Stacks, Queues, and Heaps Monday, February 18 / Tuesday, February 19

CIS 121 Data Structures and Algorithms with Java Spring Stacks, Queues, and Heaps Monday, February 18 / Tuesday, February 19 CIS Data Structures ad Algorithms with Java Sprig 09 Stacks, Queues, ad Heaps Moday, February 8 / Tuesday, February 9 Stacks ad Queues Recall the stack ad queue ADTs (abstract data types from lecture.

More information

A Comparative Study of Positive and Negative Factorials

A Comparative Study of Positive and Negative Factorials A Comparative Study of Positive ad Negative Factorials A. M. Ibrahim, A. E. Ezugwu, M. Isa Departmet of Mathematics, Ahmadu Bello Uiversity, Zaria Abstract. This paper preset a comparative study of the

More information

Assignment Problems with fuzzy costs using Ones Assignment Method

Assignment Problems with fuzzy costs using Ones Assignment Method IOSR Joural of Mathematics (IOSR-JM) e-issn: 8-8, p-issn: 9-6. Volume, Issue Ver. V (Sep. - Oct.06), PP 8-89 www.iosrjourals.org Assigmet Problems with fuzzy costs usig Oes Assigmet Method S.Vimala, S.Krisha

More information

Performance Plus Software Parameter Definitions

Performance Plus Software Parameter Definitions Performace Plus+ Software Parameter Defiitios/ Performace Plus Software Parameter Defiitios Chapma Techical Note-TG-5 paramete.doc ev-0-03 Performace Plus+ Software Parameter Defiitios/2 Backgroud ad Defiitios

More information

Evaluation scheme for Tracking in AMI

Evaluation scheme for Tracking in AMI A M I C o m m u i c a t i o A U G M E N T E D M U L T I - P A R T Y I N T E R A C T I O N http://www.amiproject.org/ Evaluatio scheme for Trackig i AMI S. Schreiber a D. Gatica-Perez b AMI WP4 Trackig:

More information

Elementary Educational Computer

Elementary Educational Computer Chapter 5 Elemetary Educatioal Computer. Geeral structure of the Elemetary Educatioal Computer (EEC) The EEC coforms to the 5 uits structure defied by vo Neuma's model (.) All uits are preseted i a simplified

More information

AN OPTIMIZATION NETWORK FOR MATRIX INVERSION

AN OPTIMIZATION NETWORK FOR MATRIX INVERSION 397 AN OPTIMIZATION NETWORK FOR MATRIX INVERSION Ju-Seog Jag, S~ Youg Lee, ad Sag-Yug Shi Korea Advaced Istitute of Sciece ad Techology, P.O. Box 150, Cheogryag, Seoul, Korea ABSTRACT Iverse matrix calculatio

More information

Description of some supervised learning algorithms

Description of some supervised learning algorithms Descriptio of some supervised learig algorithms Patrick Keekayoro patrick.keekayoro@outlook.com Statistical Cybermetrics Research Group Uiversity of Wolverhampto 1. Supervised learig Supervised machie

More information

Unsupervised Discretization Using Kernel Density Estimation

Unsupervised Discretization Using Kernel Density Estimation Usupervised Discretizatio Usig Kerel Desity Estimatio Maregle Biba, Floriaa Esposito, Stefao Ferilli, Nicola Di Mauro, Teresa M.A Basile Departmet of Computer Sciece, Uiversity of Bari Via Oraboa 4, 7025

More information

Appendix D. Controller Implementation

Appendix D. Controller Implementation COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Appedix D Cotroller Implemetatio Cotroller Implemetatios Combiatioal logic (sigle-cycle); Fiite state machie (multi-cycle, pipelied);

More information

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5

Morgan Kaufmann Publishers 26 February, COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Interface. Chapter 5 Morga Kaufma Publishers 26 February, 28 COMPUTER ORGANIZATION AND DESIGN The Hardware/Software Iterface 5 th Editio Chapter 5 Set-Associative Cache Architecture Performace Summary Whe CPU performace icreases:

More information

CMPT 125 Assignment 2 Solutions

CMPT 125 Assignment 2 Solutions CMPT 25 Assigmet 2 Solutios Questio (20 marks total) a) Let s cosider a iteger array of size 0. (0 marks, each part is 2 marks) it a[0]; I. How would you assig a poiter, called pa, to store the address

More information

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method

A New Morphological 3D Shape Decomposition: Grayscale Interframe Interpolation Method A ew Morphological 3D Shape Decompositio: Grayscale Iterframe Iterpolatio Method D.. Vizireau Politehica Uiversity Bucharest, Romaia ae@comm.pub.ro R. M. Udrea Politehica Uiversity Bucharest, Romaia mihea@comm.pub.ro

More information

Optimization for framework design of new product introduction management system Ma Ying, Wu Hongcui

Optimization for framework design of new product introduction management system Ma Ying, Wu Hongcui 2d Iteratioal Coferece o Electrical, Computer Egieerig ad Electroics (ICECEE 2015) Optimizatio for framework desig of ew product itroductio maagemet system Ma Yig, Wu Hogcui Tiaji Electroic Iformatio Vocatioal

More information

Algorithm Selection using Reinforcement Learning

Algorithm Selection using Reinforcement Learning Algorithm Selectio usig Reiforcemet Learig Michail G. Lagoudakis Departmet of Computer Sciece, Duke Uiversity, Durham, NC 2778, USA Michael L. Littma Shao Laboratory, AT&T Labs Research, Florham Park,

More information

An Efficient Algorithm for Graph Bisection of Triangularizations

An Efficient Algorithm for Graph Bisection of Triangularizations Applied Mathematical Scieces, Vol. 1, 2007, o. 25, 1203-1215 A Efficiet Algorithm for Graph Bisectio of Triagularizatios Gerold Jäger Departmet of Computer Sciece Washigto Uiversity Campus Box 1045, Oe

More information

Random Graphs and Complex Networks T

Random Graphs and Complex Networks T Radom Graphs ad Complex Networks T-79.7003 Charalampos E. Tsourakakis Aalto Uiversity Lecture 3 7 September 013 Aoucemet Homework 1 is out, due i two weeks from ow. Exercises: Probabilistic iequalities

More information

BAYESIAN WITH FULL CONDITIONAL POSTERIOR DISTRIBUTION APPROACH FOR SOLUTION OF COMPLEX MODELS. Pudji Ismartini

BAYESIAN WITH FULL CONDITIONAL POSTERIOR DISTRIBUTION APPROACH FOR SOLUTION OF COMPLEX MODELS. Pudji Ismartini Proceedig of Iteratioal Coferece O Research, Implemetatio Ad Educatio Of Mathematics Ad Scieces 014, Yogyakarta State Uiversity, 18-0 May 014 BAYESIAN WIH FULL CONDIIONAL POSERIOR DISRIBUION APPROACH FOR

More information

Performance Comparisons of PSO based Clustering

Performance Comparisons of PSO based Clustering Performace Comparisos of PSO based Clusterig Suresh Chadra Satapathy, 2 Guaidhi Pradha, 3 Sabyasachi Pattai, 4 JVR Murthy, 5 PVGD Prasad Reddy Ail Neeruoda Istitute of Techology ad Scieces, Sagivalas,Vishaapatam

More information

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP

Introduction. Nature-Inspired Computing. Terminology. Problem Types. Constraint Satisfaction Problems - CSP. Free Optimization Problem - FOP Nature-Ispired Computig Hadlig Costraits Dr. Şima Uyar September 2006 Itroductio may practical problems are costraied ot all combiatios of variable values represet valid solutios feasible solutios ifeasible

More information

Descriptive Statistics Summary Lists

Descriptive Statistics Summary Lists Chapter 209 Descriptive Statistics Summary Lists Itroductio This procedure is used to summarize cotiuous data. Large volumes of such data may be easily summarized i statistical lists of meas, couts, stadard

More information

Accuracy Improvement in Camera Calibration

Accuracy Improvement in Camera Calibration Accuracy Improvemet i Camera Calibratio FaJie L Qi Zag ad Reihard Klette CITR, Computer Sciece Departmet The Uiversity of Aucklad Tamaki Campus, Aucklad, New Zealad fli006, qza001@ec.aucklad.ac.z r.klette@aucklad.ac.z

More information

Text Feature Selection based on Feature Dispersion Degree and Feature Concentration Degree

Text Feature Selection based on Feature Dispersion Degree and Feature Concentration Degree Available olie at www.ijpe-olie.com vol. 13, o. 7, November 017, pp. 1159-1164 DOI: 10.3940/ijpe.17.07.p19.11591164 Text Feature Selectio based o Feature Dispersio Degree ad Feature Cocetratio Degree Zhifeg

More information

Text Summarization using Neural Network Theory

Text Summarization using Neural Network Theory Iteratioal Joural of Computer Systems (ISSN: 2394-065), Volume 03 Issue 07, July, 206 Available at http://www.ijcsolie.com/ Simra Kaur Jolly, Wg Cdr Ail Chopra 2 Departmet of CSE, Ligayas Uiversity, Faridabad

More information