FuzzConRI - A Fuzzy Conjunctive Rule Inducer

Size: px
Start display at page:

Download "FuzzConRI - A Fuzzy Conjunctive Rule Inducer"

Transcription

1 FuzzConRI - A Fuzzy Conjunctive Rule Inducer Jacobus van Zyl and Ian Cloete School of Information Technology, International University in Germany, Bruchsal, Germany {jacobus.vanzyl,ian.cloete}@i-u.de Abstract. A variety of methods exist for inductive learning of classification rules using crisp sets. In this paper we illustrate an inductive learner that uses fuzzy sets, where the membership functions of the linguistic terms are given in advance. We also show how the induction of conjunctive rules fit into a fuzzy set covering framework (FuzzyBexa) that we introduced before. 1 Introduction Rule inducers are an important category of machine learning algorithms. Typically they obtain propositional expressions derived from a set of training instances. A multitude rule learning algorithms were proposed, and especially the separate-and-conquer approach to rule learning has been very popular [1]. These algorithms are all based on crisp set covering ideas. Fuzzy sets are a generalization of crisp sets [2] providing increased expressive power and comprehensibility. There have been many attempts to induce fuzzy concept representations from data. These include fuzzy neural networks [3], fuzzy decision trees [4], genetic algorithms [5], grid methods [6], and clustering [7]. These methods, however, are primarily focused on the parameter identification, i.e. finding suitable parameters for a fuzzy inference systems, such as membership functions for the fuzzy sets. Little work has been done thus far on rule induction systems that perform structure identification, i.e. learning incomplete rules that do not contain all the input domain variables in the antecedent. Some efforts have been made by Wang et al who propose an induction method for learning modular fuzzy rules [8] which is based on the crisp rule learner PRISM [9]. Castro et al showed that an assumption-based truth maintenance system (ATMS) can be used to build a fuzzy inductive learner. This algorithm finds the minimal node in an ATMS and uses it to propose fuzzy classification rules [10]. In this paper we present FuzzConRI, a novel induction algorithm capable of inducing incomplete fuzzy classification rules from data. We also show that this algorithm can fit into the fuzzy set covering framework FuzzyBexa [11]. 2 FuzzConRI FuzzConRI is a fuzzy rule induction algorithm based on the CN2 approach [12, 13]. The algorithm is shown in Table 1. It consists of two layers, an upper

2 Table 1. The FuzzConRI algorithm FuzzConRI Input: Set of training instances T, Set of concepts to learn Concepts Output: A rule set describing the concepts Set the rule set to empty FOR EACH concept Concepts P = {i T µ concept(i) α c} N = T P WHILE P is not empty, and more rules can be found DO antecedent=findbestantecedent(p, N) If a suitable antecedent is found, augment the rule set with IF antecedent THEN concept Remove the positive instances covered by the added rule Return the rule set FindBestAntecedent Input: Set of positive instances, Set of negative instances Output: Antecedent that covers the positive instances best Let ST AR contain the antecedent T RUE Let BEST ANT be nil Let T ERMS contain all possible terms While ST AR is not empty NEW ST AR = {x y x ST AR, y T ERMS} NEW ST AR = NEW ST AR ST AR For each antecedent A in NEW ST AR If A is better than BEST ANT according to an evaluation function, then Replace the current best conjunction with the new one Remove all antecedents that cover only positive instances Retain a user defined number of best elements in NEW ST AR ST AR = NEW ST AR Return BEST ANT layer implementing a set cover approach to rule induction, and a lower layer for inducing one rule. The upper layer receives a set of training instances T, and a set of concepts Concepts. It starts by initialising the rule set to the empty set, and then considers each concept concept Concepts, one at a time. The training set T is split into two parts, a set of fuzzy instances P that belong to the concept and a set of fuzzy instances N that do not belong to the concept. The sets P and N are called the set of positive and negative instances, respectively. In contrast to a crisp instance, a fuzzy instance belongs to all linguistic terms (attribute values) to a certain degree. Crisp instances can belong to only one attribute value per attribute at a time. If a concept such as Temperature is hot must be learned, the user specifies a threshold membership α c at which the temperature is considered hot. The procedure FindBestAntecedent is then repeatedly called to induce a conjunction that covers a set of instances in T.

3 The instances in P covered by the conjunction are removed from P and the procedure repeated until P is empty. The rules induced by FuzzConRI are of the form: IF [T emperature is mild hot] [W ind is calm] THEN Sport is tennis where T emperature, W ind and Sport are linguistic variables, and mild, hot, calm and tennis linguistic terms. Here the difference between fuzzy and crisp rule induction becomes clear. The above rule will cover no instances in the crisp case since no instance can have both T emperature = mild and T emperature = hot at the same time. This, however, is perfectly possible in the fuzzy case. The procedure FindAntecedent receives a set of positive and a set of negative instances and returns the antecedent that best covers the positive instances while attempting not to cover any negative instances. It starts by initialising the set ST AR with the mgc (most general conjunction). This conjunction should cover all possible instances, and thereby also the training set. The best conjunction found during the search is stored in the variable BEST ANT. A new ST AR is obtained by forming the conjunction of all possible single linguistic term descriptions with each conjunction in ST AR, and the removing the conjunctions that were not changed, e.g. the conjunction of [T emperature = hot][w ind = calm] with the description [W ind = hot] does not change the conjunction, and is removed. Here we left the symbol out for brevity s sake. After the new conjunctions are formed, they are evaluated according to an evaluation function. This evaluation functions plays a crucial role to guide the search for good rule antecedents. A rule evaluation function that performs very well, especially in the presence of noise, is the fuzzy Accuracy function, A(r) = X P (c) (X N (c) (1) where c is a conjunction, and X S (c) is its extension in the set S, i.e. all the instances that are matched by c. The set cardinality in the fuzzy case, also called the sigma count of the set [14], is computed as, X S (c) = µ c (i) (2) i X S(c) We consider an instance to match a fuzzy conjunction when its membership to the conjunction is non-zero. If necessary, an α-cut can also be applied to the fuzzy sets in the input domain. In this case the value of the α-cut is specified by α a. The fuzzy and operators can be implemented by any t-norm and t- conorm functions, such as minimum and maximum for example. For an overview of possible fuzzy conjunction evaluation functions for iterated concept learning refer to [15, 16]. After the conjunctions are evaluated, they are compared to the best conjunction found thus far, and if a conjunction obtains a higher score than the best conjunction it replaces the best conjunction. Finally, the conjunctions that cover no positive instances are removed from the search, since they cannot be

4 Table 2. FuzzyBexa s top layer routines. CoverConcepts Input: Set of training instances, Set of concepts to learn Output: A rule set describing the concepts Set the current rule set to empty For each concept Find the set of positive and negative instances While there are positive instances Find the best conjunction If a suitable conjunction is found, augment the rule set with IF conjunction THEN concept Remove the positive instances covered by the added rule Return the rule set FindBestConjunction Input: Set of positive instances, Set of negative instances Output: Conjunction that covers the positive instances best Set the current best conjunction to empty Add the most general conjunction to the current conjunctions While more (useful) conjunctions can be found Generate more specializations For each new conjunction If the conjunction is better than the current best conjunction according to an evaluation function, Replace the current best conjunction with the new one Remove conjunctions that cover no negative instances, or that cannot be improved beyond the performance of the best conjunction Retain only the beamwidth best conjunctions If the evaluation function value for the best conjunction is the same or worse than that of the complete training set, return: no conjunction found, otherwise, return the best conjunction found. improved upon. FuzzConRI can implement a beam search by retaining more than one conjunction for further specialization in the new ST AR. The set ST AR will eventually become empty when no more new conjunctions can be formed such that they do not replace older conjunctions. The best conjunction found during the search process is then returned. 3 FuzzConRI and FuzzyBexa The BEXA framework provided a method for comparison of several crisp rule induction algorithms [17]. We recently extended this framework to the fuzzy case [11]. FuzzyBexa consists of three layers. The top layer implements a fuzzy set

5 Table 3. FuzzConRI as specialization model. GenerateSpecializations Input: Set of conjunctions to specialize Output: Set of specializations Let specializtions be the empty set. For each conjunction c For each linguistic term L Add the conjunction cnew = c L to specializations Remove all duplicate conjunctions from specializations Return specializations covering approach, the middle layer implements several heuristics for guiding the search process, and the bottom layer implements conjunction refinement. The framework uses the description language FuzzyVL 1 which allows internal disjunction. The method of conjunction refinement is implemented by the specialization operator. One possible specialization operator is exclusion, where conjunctions are specialized by removing (excluding) a linguistic term from the description. By changing FuzzyBexa s specialization operator from exclusion to append, FuzzConRI behaviour is implemented by the framework. FuzzyBexa s top layer routines are shown in Table 2, and Table 3 shows FuzzConRI as a specialization model within the framework. FuzzyBexa s top layer routines implement many different heuristics that can be shared between different specialization models. As such, the extra work needed to implement FuzzConRI s behaviour is indeed not much. For an in-depth discussion of FuzzyBexa s induction strategies, search heuristics, efficiency measures, rule pre-pruning methods and stopping criteria refer to reference [11]. The rules induced by FuzzConRI are unordered. Recently, FuzzyBexaII was introduced that extended FuzzyBexa to allow the use of a simultaneous concept learning strategy [18]. 4 Experimental Results In this section we provide an empirical evaluation of different parameters for FuzzConRI. The data sets are available from the UCI machine learning repository [19]. Unfortunately there is currently a lack of publicly available originally fuzzy data sets. By originally fuzzy we mean that either the fuzzy membership functions are known or provided by experts, or that instances are characterized by their membership degrees to linguistic terms. We obtain a fuzzy data set by a process of data fuzzification. Crisp nominal attributes are converted to linguistic variables with crisp membership functions. Fuzzy membership functions are obtained for numerical attributes by first clustering the instance values and then placing bell-shaped membership functions of the form published in [20] at the cluster centres. However, here we have to stress that our methods are independent of the method of parameter identification (membership function

6 extraction), and that better methods for membership function extraction will of course improve the overall system performance. Figure 1 shows results obtained by FuzzConRI from 10-fold cross validation on the Iris and Bodyfat data sets. Figure 1(a) shows the influence of changing the value of α a, i.e. applying an α-cut at varying membership degrees. Here we use the same value for α a during rule induction and test set evaluation (inferencing). Results are shown for the classification accuracy, the rule set complexity, and the amount of search performed to induce the rule set. Rule set complexity is measured by the number of linguistic variables used in the rule set. FuzzConRI s classification performance on the Iris data set seems to be very insensitive to α a. Only extreme values of α a results in a degraded classification accuracy. A similar result is obtained for the rule set complexity. However, the number of conjunctions generated during the induction of the rule set is more sensitive to extreme values for α a, with 2.5 times as many needed for α a = 0.95 than for 0 α a Figure 1(b) shows the effect of the beam width on the induction process. Here α a was kept constant and the beam width was set to values 1, 2, 3, 5, 10, 15, 25, and 50. As is clear from the figure, the beam width had no effect on the classification accuracy or the rule set complexity. This implies that FuzzConRI already obtain the best solution without employing a beam search. A further observation from the figure is that beam widths of 10 and more do not result in the generation of more conjunctions. The reason for that is FuzzyBexa s efficiency measures. FuzzyBexa removes conjunctions that cannot result in good concept descriptions and only concentrate on interesting parts of the search space. Thus a very efficient search of the space of conjunctions is performed, and a large beam width is not required to obtain good results. Figure 1(c) shows the effect of α a for the Bodyfat data. Here a good value for α a is clearly important, and the data set shows sensitivity to α a for all measured parameters. The optimum value for classification performance is obtained for α a 0.65, and the least complex rule set is obtained at α a An interesting trend is the correlation between rule set complexity and the size of the search space examined. This correlation is due to the fact the rule set complexity is correlated with the number of rules in the rule set, and for each rule the hypothesis space is searched. Thus, the number of rules per rule set is directly related to the amount of search required, as can clearly be seen in Figure 1(c). Note that in Figure 1(c) the amount of search shown on the graph is multiplied with a scale factor of 0.1. It is not clear how to automatically choose the value for α a in general. The optimum value for α a might depend on the method of membership function extraction. Our fuzzification method placed bell shaped function on the cluster centres. For the Bodyfat data, values of α a greater than 0.6 resulted in little overlap of the membership functions. This had a detrimental effect on performance. When there is very little overlap, the fuzzy membership functions revert back to crisp -like behaviour, with no two fuzzy sets from the same variable being non-zero at the same time. Thus, in Figure 1(c) the beneficial effect of the fuzzy

7 Influence of α a for the Iris Data Number of Conjunctions Generated 120 Influence of the beam width on the Iris data Number of Conjunctions Generated α a (a) Influence of α a for the Iris data Influence of α a on the Bodyfat data Number of Conjunctions Generated (Scale 0.1) Beam Width (b) Influence of the beam width for the Iris data Influence of the beam width on the Bodyfat data Number of Conjunctions Generated (Scale 0.01) α a (c) Influence of α a for the Bodyfat data Beam Width (d) Influence of the beam width for the Bodyfat data Fig. 1. Results the Iris and Bodyfat data induction method is clearly shown in the rule complexity and search requirement behaviour. For values of α a greater then 0.65 the classification performance also decreased steadily. The reason for the behaviour of the classification performance for α a < 0.65 is not clear. Figure 1(d) shows the effect of an increased beam width on the Bodyfat data. Similar to the Iris data, a near optimum classification result is obtained with no beam search. A marginal increase in classification performance is obtained for a beam width greater than 10. Figure 2(a) and (b) shows results for the Colic data. Classification accuracy is relatively insensitive to α a, while the search requirement and rule set complexity shows a complex relationship to α a. The beam width again had relatively little effect, with only slightly less complex rule sets obtained for greater beam widths. A further observation is that the increase in search requirement is rather linear with an increase in beam width

8 Influence of the α a on the Colic data Number of Conjunctions Generated (Scale 0.001) Influence of the beam width on the Colic data Number of Conjunctions Generated (Scale 0.001) α a (a) Influence of α a for the Colic data Influence of α a on the Lymph data Beam Width (b) Influence of the beam width for the Colic data Influence of the beam width on the Lymph data Number of Conjunctions Generated (Scale 0.1) α a (c) Influence of α a for the Lymph data 10 Number of Conjunctions Generated (Scale 0.001) Beam Width (d) Influence of the beam width for the Lymph data Fig. 2. Results the Colic and Lymph data than exponential. In fact, from Figures 1(a) and (c) we may deduce that the trend of the search requirement is initially linear and eventually reach a plateau. Figure 2(c) shows the relative insensitivity to α a for the Lymph data. Figure 2(d) shows that an increased beam width can in some cases have a positive effect on classification performance and rule set complexity. 5 Conclusion In this paper we presented FuzzConRI, a novel algorithm that can induce fuzzy conjunctive classification rules from data. The algorithm is based on the approach by CN2, but differ from crisp inductive algorithms in that fuzzy sets are used in the description languages. Fuzzy instances belong to all linguistic terms in the problem to a given degree, and therefore rules may have antecedents such

9 as T emperature = cold mild. We have also reviewed FuzzyBexa, a set covering framework for rule induction. We have shown that FuzzConRI fits in this framework and how to insert FuzzConRI as a specialization model in Fuzzy- Bexa. Finally we provided empirical results of FuzzConRI s performance on four data sets for different induction parameters. From the experiments we conclude that FuzzConRI can induce very good rule descriptions even when using no beam search. The experiments have further shown that in a few cases Fuzz- ConRI is sensitive to the value of α a, the α-cut applied to the input domain membership degrees. However, performance is never completely degraded, and in most cases FuzzConRI s classification performance is rather insensitive to α a. References 1. Fürnkranz, J.: Separate-and-conquer rule learning. Artificial Intelligence Review 13 (1999) Cox, E.: The Fuzzy Systems Handbook. 2 edn. Academic Press, London (1998) 3. Kasabov, N.K.: On-line learning, reasoning, rule extraction and aggregation in locally optimized evolving fuzzy neural networks. Neurocomputing (2001) Singal, P.K., Mitra, S., Pal, S.K.: Incorporation of fuzziness in id3 and generation of network architecture. Neural Computing and Applications 10 (2001) Cordón, O., Herrera, F., Hoffmann, F., Magdalena, L.: Genetic Fuzzy Systems - Evolutionary Tuning and Learning of Fuzzy Knowledge Bases. World Scientific, Singapore (2001) 6. Wang, L., Mendel, J.M.: Generating fuzzy rules by learning from examples. IEEE Trans. on Systems, Man and Cybernetics 22 (1992) Gedeon, T.D., Kuo, H., Wong, P.M.: Rule extraction using fuzzy clustering for a sedimentary rock data set. International Journal of Fuzzy Systems 4 (2002) Wang, C.H., Liu, J.F., Hong, T.P., Tseng, S.S.: A fuzzy inductive learning strategy for modular rules. Fuzzy Sets and Systems 103 (1999) Cendrowska, J.: PRISM: An algorithm for inducing modular rules. International Journal of Man-Machines Studies 27 (1987) Castro, J.L., Zurita, J.M.: An inductive learning algorithm in fuzzy systems. Fuzzy Sets and Systems 89 (1997) Cloete, I., van Zyl, J.: Fuzzy rule induction in a set covering framework. (2004) (Submitted). 12. Clark, P., Boswell, R.: Rule induction with CN2: Some recent improvements. In: Proceedings of the Sixth European Working Session on Learning. (1991) van Zyl, J., Cloete, I.: An inductive algorithm for learning conjunctive fuzzy rules. In: International Conference on Machine Learning and Cybernetics, Shanghai, China (2004) 14. Klir, G.J., Yuan, B.: Fuzzy Sets and Fuzzy Logic. Prentice Hall, Upper Saddle River, NJ (1995) 15. van Zyl, J., Cloete, I.: Heuristic functions for learning fuzzy conjunctive rules. In: IEEE International Conference on Systems, Man and Cybernetics, The Hague, The Netherlands (2004) 16. Cloete, I., van Zyl, J.: Evaluation function guided search for fuzzy set covering. In: IEEE International Conference on Fuzzy Systems, Budapest, Hungary (2004)

10 17. Theron, H., Cloete, I.: BEXA: A covering algorithm for learning propositional concept descriptions. Machine Learning 24 (1996) van Zyl, J., Cloete, I.: Simultaneous concept learning of fuzzy rules. In: Proceedings of the 15th European Conference on Machine Learning, Pisa, Italy (2004) 19. Blake, C.L., Merz, C.J.: UCI repository of machine learning databases (1998) URL: mlearn/. 20. Surmann, H.: Learning a fuzzy rule based knowledge representation. In: Proc. of 2. ICSC Symp. on Neural Computation, Berlin (2000)

Simultaneous Concept Learning of Fuzzy Rules

Simultaneous Concept Learning of Fuzzy Rules Simultaneous Concept Learning of Fuzzy Rules JacobusvanZylandIanCloete School of Information Technology, International University in Germany, 76646 Bruchsal, Germany Abstract. FuzzyBexa was the first algorithm

More information

Fuzzy Partitioning with FID3.1

Fuzzy Partitioning with FID3.1 Fuzzy Partitioning with FID3.1 Cezary Z. Janikow Dept. of Mathematics and Computer Science University of Missouri St. Louis St. Louis, Missouri 63121 janikow@umsl.edu Maciej Fajfer Institute of Computing

More information

Introduction to Fuzzy Logic. IJCAI2018 Tutorial

Introduction to Fuzzy Logic. IJCAI2018 Tutorial Introduction to Fuzzy Logic IJCAI2018 Tutorial 1 Crisp set vs. Fuzzy set A traditional crisp set A fuzzy set 2 Crisp set vs. Fuzzy set 3 Crisp Logic Example I Crisp logic is concerned with absolutes-true

More information

Genetic Tuning for Improving Wang and Mendel s Fuzzy Database

Genetic Tuning for Improving Wang and Mendel s Fuzzy Database Proceedings of the 2009 IEEE International Conference on Systems, Man, and Cybernetics San Antonio, TX, USA - October 2009 Genetic Tuning for Improving Wang and Mendel s Fuzzy Database E. R. R. Kato, O.

More information

Application of Or-based Rule Antecedent Fuzzy Neural Networks to Iris Data Classification Problem

Application of Or-based Rule Antecedent Fuzzy Neural Networks to Iris Data Classification Problem Vol.1 (DTA 016, pp.17-1 http://dx.doi.org/10.157/astl.016.1.03 Application of Or-based Rule Antecedent Fuzzy eural etworks to Iris Data Classification roblem Chang-Wook Han Department of Electrical Engineering,

More information

FEATURE EXTRACTION USING FUZZY RULE BASED SYSTEM

FEATURE EXTRACTION USING FUZZY RULE BASED SYSTEM International Journal of Computer Science and Applications, Vol. 5, No. 3, pp 1-8 Technomathematics Research Foundation FEATURE EXTRACTION USING FUZZY RULE BASED SYSTEM NARENDRA S. CHAUDHARI and AVISHEK

More information

Learning Fuzzy Rules Using Ant Colony Optimization Algorithms 1

Learning Fuzzy Rules Using Ant Colony Optimization Algorithms 1 Learning Fuzzy Rules Using Ant Colony Optimization Algorithms 1 Jorge Casillas, Oscar Cordón, Francisco Herrera Department of Computer Science and Artificial Intelligence, University of Granada, E-18071

More information

What is all the Fuzz about?

What is all the Fuzz about? What is all the Fuzz about? Fuzzy Systems CPSC 433 Christian Jacob Dept. of Computer Science Dept. of Biochemistry & Molecular Biology University of Calgary Fuzzy Systems in Knowledge Engineering Fuzzy

More information

Exploring Gaussian and Triangular Primary Membership Functions in Non-Stationary Fuzzy Sets

Exploring Gaussian and Triangular Primary Membership Functions in Non-Stationary Fuzzy Sets Exploring Gaussian and Triangular Primary Membership Functions in Non-Stationary Fuzzy Sets S. Musikasuwan and J.M. Garibaldi Automated Scheduling, Optimisation and Planning Group University of Nottingham,

More information

A New Approach for Handling the Iris Data Classification Problem

A New Approach for Handling the Iris Data Classification Problem International Journal of Applied Science and Engineering 2005. 3, : 37-49 A New Approach for Handling the Iris Data Classification Problem Shyi-Ming Chen a and Yao-De Fang b a Department of Computer Science

More information

FUZZY BOOLEAN ALGEBRAS AND LUKASIEWICZ LOGIC. Angel Garrido

FUZZY BOOLEAN ALGEBRAS AND LUKASIEWICZ LOGIC. Angel Garrido Acta Universitatis Apulensis ISSN: 1582-5329 No. 22/2010 pp. 101-111 FUZZY BOOLEAN ALGEBRAS AND LUKASIEWICZ LOGIC Angel Garrido Abstract. In this paper, we analyze the more adequate tools to solve many

More information

A Multi-objective Genetic-based Method for Design Fuzzy Classification Systems

A Multi-objective Genetic-based Method for Design Fuzzy Classification Systems IJCSNS International Journal of Computer Science and Network Security, VOL.6 No.8A, August 6 A Multi-obective Genetic-based Method for Design Fuzzy Classification Systems Chen Ji-lin, Hou Yuan-long, Xing

More information

Enhancing Forecasting Performance of Naïve-Bayes Classifiers with Discretization Techniques

Enhancing Forecasting Performance of Naïve-Bayes Classifiers with Discretization Techniques 24 Enhancing Forecasting Performance of Naïve-Bayes Classifiers with Discretization Techniques Enhancing Forecasting Performance of Naïve-Bayes Classifiers with Discretization Techniques Ruxandra PETRE

More information

Identifying fuzzy systems from numerical data with Xfuzzy

Identifying fuzzy systems from numerical data with Xfuzzy Identifying fuzzy systems from numerical data with Xfuzzy I. Baturone Instituto de Microelectrónica de Sevilla, (IMSE-CNM), y Dept. Electrónica y Electromagnetismo, Univ. de Sevilla (Spain) lumi@imse.cnm.es

More information

Improving the Wang and Mendel s Fuzzy Rule Learning Method by Inducing Cooperation Among Rules 1

Improving the Wang and Mendel s Fuzzy Rule Learning Method by Inducing Cooperation Among Rules 1 Improving the Wang and Mendel s Fuzzy Rule Learning Method by Inducing Cooperation Among Rules 1 J. Casillas DECSAI, University of Granada 18071 Granada, Spain casillas@decsai.ugr.es O. Cordón DECSAI,

More information

fuzzylite a fuzzy logic control library in C++

fuzzylite a fuzzy logic control library in C++ fuzzylite a fuzzy logic control library in C++ Juan Rada-Vilela jcrada@fuzzylite.com Abstract Fuzzy Logic Controllers (FLCs) are software components found nowadays within well-known home appliances such

More information

Multiobjective Formulations of Fuzzy Rule-Based Classification System Design

Multiobjective Formulations of Fuzzy Rule-Based Classification System Design Multiobjective Formulations of Fuzzy Rule-Based Classification System Design Hisao Ishibuchi and Yusuke Nojima Graduate School of Engineering, Osaka Prefecture University, - Gakuen-cho, Sakai, Osaka 599-853,

More information

Weighting and selection of features.

Weighting and selection of features. Intelligent Information Systems VIII Proceedings of the Workshop held in Ustroń, Poland, June 14-18, 1999 Weighting and selection of features. Włodzisław Duch and Karol Grudziński Department of Computer

More information

European Journal of Science and Engineering Vol. 1, Issue 1, 2013 ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM IDENTIFICATION OF AN INDUCTION MOTOR

European Journal of Science and Engineering Vol. 1, Issue 1, 2013 ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM IDENTIFICATION OF AN INDUCTION MOTOR ADAPTIVE NEURO-FUZZY INFERENCE SYSTEM IDENTIFICATION OF AN INDUCTION MOTOR Ahmed A. M. Emam College of Engineering Karrary University SUDAN ahmedimam1965@yahoo.co.in Eisa Bashier M. Tayeb College of Engineering

More information

Han Liu, Alexander Gegov & Mihaela Cocea

Han Liu, Alexander Gegov & Mihaela Cocea Rule-based systems: a granular computing perspective Han Liu, Alexander Gegov & Mihaela Cocea Granular Computing ISSN 2364-4966 Granul. Comput. DOI 10.1007/s41066-016-0021-6 1 23 Your article is published

More information

Implementation of Magnified Edge Detection using Fuzzy-Canny Logic

Implementation of Magnified Edge Detection using Fuzzy-Canny Logic Implementation of Magnified Edge Detection using Fuzzy-Canny Logic Hitesh Kapoor Department of Computer Science & Engineering Doon Valley Institute of Engineering & Technology Karnal, India E-mail: hitesh0409@gmail.com

More information

A Comparison of Global and Local Probabilistic Approximations in Mining Data with Many Missing Attribute Values

A Comparison of Global and Local Probabilistic Approximations in Mining Data with Many Missing Attribute Values A Comparison of Global and Local Probabilistic Approximations in Mining Data with Many Missing Attribute Values Patrick G. Clark Department of Electrical Eng. and Computer Sci. University of Kansas Lawrence,

More information

Genetic Programming for Data Classification: Partitioning the Search Space

Genetic Programming for Data Classification: Partitioning the Search Space Genetic Programming for Data Classification: Partitioning the Search Space Jeroen Eggermont jeggermo@liacs.nl Joost N. Kok joost@liacs.nl Walter A. Kosters kosters@liacs.nl ABSTRACT When Genetic Programming

More information

Introduction 3 Fuzzy Inference. Aleksandar Rakić Contents

Introduction 3 Fuzzy Inference. Aleksandar Rakić Contents Beograd ETF Fuzzy logic Introduction 3 Fuzzy Inference Aleksandar Rakić rakic@etf.rs Contents Mamdani Fuzzy Inference Fuzzification of the input variables Rule evaluation Aggregation of rules output Defuzzification

More information

ARTIFICIAL INTELLIGENCE. Uncertainty: fuzzy systems

ARTIFICIAL INTELLIGENCE. Uncertainty: fuzzy systems INFOB2KI 2017-2018 Utrecht University The Netherlands ARTIFICIAL INTELLIGENCE Uncertainty: fuzzy systems Lecturer: Silja Renooij These slides are part of the INFOB2KI Course Notes available from www.cs.uu.nl/docs/vakken/b2ki/schema.html

More information

Inducing Fuzzy Decision Trees in Non-Deterministic Domains using CHAID

Inducing Fuzzy Decision Trees in Non-Deterministic Domains using CHAID Inducing Fuzzy Decision Trees in Non-Deterministic Domains using CHAID Jay Fowdar, Zuhair Bandar, Keeley Crockett The Intelligent Systems Group Department of Computing and Mathematics John Dalton Building

More information

Evolution of Recurrent Fuzzy Controllers

Evolution of Recurrent Fuzzy Controllers Evolution of Recurrent Fuzzy Controllers Carlos Kavka, Patricia Roggero and Javier Apolloni LIDIC Departamento de Informática Universidad Nacional de San Luis Ejército de los Andes 950 D5700HHW - San Luis

More information

Matrix Inference in Fuzzy Decision Trees

Matrix Inference in Fuzzy Decision Trees Matrix Inference in Fuzzy Decision Trees Santiago Aja-Fernández LPI, ETSIT Telecomunicación University of Valladolid, Spain sanaja@tel.uva.es Carlos Alberola-López LPI, ETSIT Telecomunicación University

More information

MODELING FOR RESIDUAL STRESS, SURFACE ROUGHNESS AND TOOL WEAR USING AN ADAPTIVE NEURO FUZZY INFERENCE SYSTEM

MODELING FOR RESIDUAL STRESS, SURFACE ROUGHNESS AND TOOL WEAR USING AN ADAPTIVE NEURO FUZZY INFERENCE SYSTEM CHAPTER-7 MODELING FOR RESIDUAL STRESS, SURFACE ROUGHNESS AND TOOL WEAR USING AN ADAPTIVE NEURO FUZZY INFERENCE SYSTEM 7.1 Introduction To improve the overall efficiency of turning, it is necessary to

More information

Equi-sized, Homogeneous Partitioning

Equi-sized, Homogeneous Partitioning Equi-sized, Homogeneous Partitioning Frank Klawonn and Frank Höppner 2 Department of Computer Science University of Applied Sciences Braunschweig /Wolfenbüttel Salzdahlumer Str 46/48 38302 Wolfenbüttel,

More information

Subsethood-based Fuzzy Modelling and Classification

Subsethood-based Fuzzy Modelling and Classification Subsethood-based Fuzzy Modelling and Classification Khairul Anwar Rasmani Centre for Intelligent Systems and their Applications, School of Informatics, The University of Edinburgh K.A.Rasmani@ sms.ed.ac.u

More information

Data Mining. 3.3 Rule-Based Classification. Fall Instructor: Dr. Masoud Yaghini. Rule-Based Classification

Data Mining. 3.3 Rule-Based Classification. Fall Instructor: Dr. Masoud Yaghini. Rule-Based Classification Data Mining 3.3 Fall 2008 Instructor: Dr. Masoud Yaghini Outline Using IF-THEN Rules for Classification Rules With Exceptions Rule Extraction from a Decision Tree 1R Algorithm Sequential Covering Algorithms

More information

DEVELOPMENT OF NEURAL NETWORK TRAINING METHODOLOGY FOR MODELING NONLINEAR SYSTEMS WITH APPLICATION TO THE PREDICTION OF THE REFRACTIVE INDEX

DEVELOPMENT OF NEURAL NETWORK TRAINING METHODOLOGY FOR MODELING NONLINEAR SYSTEMS WITH APPLICATION TO THE PREDICTION OF THE REFRACTIVE INDEX DEVELOPMENT OF NEURAL NETWORK TRAINING METHODOLOGY FOR MODELING NONLINEAR SYSTEMS WITH APPLICATION TO THE PREDICTION OF THE REFRACTIVE INDEX THESIS CHONDRODIMA EVANGELIA Supervisor: Dr. Alex Alexandridis,

More information

An Information-Theoretic Approach to the Prepruning of Classification Rules

An Information-Theoretic Approach to the Prepruning of Classification Rules An Information-Theoretic Approach to the Prepruning of Classification Rules Max Bramer University of Portsmouth, Portsmouth, UK Abstract: Keywords: The automatic induction of classification rules from

More information

The Application of K-medoids and PAM to the Clustering of Rules

The Application of K-medoids and PAM to the Clustering of Rules The Application of K-medoids and PAM to the Clustering of Rules A. P. Reynolds, G. Richards, and V. J. Rayward-Smith School of Computing Sciences, University of East Anglia, Norwich Abstract. Earlier research

More information

Identification of Vehicle Class and Speed for Mixed Sensor Technology using Fuzzy- Neural & Genetic Algorithm : A Design Approach

Identification of Vehicle Class and Speed for Mixed Sensor Technology using Fuzzy- Neural & Genetic Algorithm : A Design Approach Identification of Vehicle Class and Speed for Mixed Sensor Technology using Fuzzy- Neural & Genetic Algorithm : A Design Approach Prashant Sharma, Research Scholar, GHRCE, Nagpur, India, Dr. Preeti Bajaj,

More information

Nearest Cluster Classifier

Nearest Cluster Classifier Nearest Cluster Classifier Hamid Parvin, Moslem Mohamadi, Sajad Parvin, Zahra Rezaei, and Behrouz Minaei Nourabad Mamasani Branch, Islamic Azad University, Nourabad Mamasani, Iran hamidparvin@mamasaniiau.ac.ir,

More information

Lecture 5 Fuzzy expert systems: Fuzzy inference Mamdani fuzzy inference Sugeno fuzzy inference Case study Summary

Lecture 5 Fuzzy expert systems: Fuzzy inference Mamdani fuzzy inference Sugeno fuzzy inference Case study Summary Lecture 5 Fuzzy expert systems: Fuzzy inference Mamdani fuzzy inference Sugeno fuzzy inference Case study Summary Negnevitsky, Pearson Education, 25 Fuzzy inference The most commonly used fuzzy inference

More information

Speed regulation in fan rotation using fuzzy inference system

Speed regulation in fan rotation using fuzzy inference system 58 Scientific Journal of Maritime Research 29 (2015) 58-63 Faculty of Maritime Studies Rijeka, 2015 Multidisciplinary SCIENTIFIC JOURNAL OF MARITIME RESEARCH Multidisciplinarni znanstveni časopis POMORSTVO

More information

Efficient Pairwise Classification

Efficient Pairwise Classification Efficient Pairwise Classification Sang-Hyeun Park and Johannes Fürnkranz TU Darmstadt, Knowledge Engineering Group, D-64289 Darmstadt, Germany Abstract. Pairwise classification is a class binarization

More information

Derivation of Relational Fuzzy Classification Rules Using Evolutionary Computation

Derivation of Relational Fuzzy Classification Rules Using Evolutionary Computation Derivation of Relational Fuzzy Classification Rules Using Evolutionary Computation Vahab Akbarzadeh Alireza Sadeghian Marcus V. dos Santos Abstract An evolutionary system for derivation of fuzzy classification

More information

i-miner: A Web Usage Mining Framework Using Neuro-Genetic-Fuzzy Approach

i-miner: A Web Usage Mining Framework Using Neuro-Genetic-Fuzzy Approach i-miner: A Web Usage Mining Framework Using Neuro-Genetic-Fuzzy Approach Ajith Abraham and Xiaozhe Wang* Department of Computer Science, Oklahoma State University, 700 N Greenwood Avenue, Tulsa, OK 74106-0700,

More information

Keywords - Fuzzy rule-based systems, clustering, system design

Keywords - Fuzzy rule-based systems, clustering, system design CHAPTER 7 Application of Fuzzy Rule Base Design Method Peter Grabusts In many classification tasks the final goal is usually to determine classes of objects. The final goal of fuzzy clustering is also

More information

Estimating Missing Attribute Values Using Dynamically-Ordered Attribute Trees

Estimating Missing Attribute Values Using Dynamically-Ordered Attribute Trees Estimating Missing Attribute Values Using Dynamically-Ordered Attribute Trees Jing Wang Computer Science Department, The University of Iowa jing-wang-1@uiowa.edu W. Nick Street Management Sciences Department,

More information

Fuzzy set approximation using polar co-ordinates and linguistic term shifting

Fuzzy set approximation using polar co-ordinates and linguistic term shifting Johanyák, Z. C., Kovács, S.: Fuzzy set approximation using polar co-ordinates and linguistic term shifting, SAMI 2006, 4 rd Slovakian-Hungarian Joint Symposium on Applied Machine Intelligence, Herl'any,

More information

COSC 6397 Big Data Analytics. Fuzzy Clustering. Some slides based on a lecture by Prof. Shishir Shah. Edgar Gabriel Spring 2015.

COSC 6397 Big Data Analytics. Fuzzy Clustering. Some slides based on a lecture by Prof. Shishir Shah. Edgar Gabriel Spring 2015. COSC 6397 Big Data Analytics Fuzzy Clustering Some slides based on a lecture by Prof. Shishir Shah Edgar Gabriel Spring 215 Clustering Clustering is a technique for finding similarity groups in data, called

More information

Flexibility and Robustness of Hierarchical Fuzzy Signature Structures with Perturbed Input Data

Flexibility and Robustness of Hierarchical Fuzzy Signature Structures with Perturbed Input Data Flexibility and Robustness of Hierarchical Fuzzy Signature Structures with Perturbed Input Data B. Sumudu U. Mendis Department of Computer Science The Australian National University Canberra, ACT 0200,

More information

CHAPTER 4 FUZZY LOGIC, K-MEANS, FUZZY C-MEANS AND BAYESIAN METHODS

CHAPTER 4 FUZZY LOGIC, K-MEANS, FUZZY C-MEANS AND BAYESIAN METHODS CHAPTER 4 FUZZY LOGIC, K-MEANS, FUZZY C-MEANS AND BAYESIAN METHODS 4.1. INTRODUCTION This chapter includes implementation and testing of the student s academic performance evaluation to achieve the objective(s)

More information

This article was downloaded by: [National Chiao Tung University 國立交通大學 ]

This article was downloaded by: [National Chiao Tung University 國立交通大學 ] This article was downloaded by: [National Chiao Tung University 國立交通大學 ] On: 27 April 2014, At: 22:53 Publisher: Taylor & Francis Informa Ltd Registered in England and Wales Registered Number: 1072954

More information

A New Fuzzy Neural System with Applications

A New Fuzzy Neural System with Applications A New Fuzzy Neural System with Applications Yuanyuan Chai 1, Jun Chen 1 and Wei Luo 1 1-China Defense Science and Technology Information Center -Network Center Fucheng Road 26#, Haidian district, Beijing

More information

Swarm Based Fuzzy Clustering with Partition Validity

Swarm Based Fuzzy Clustering with Partition Validity Swarm Based Fuzzy Clustering with Partition Validity Lawrence O. Hall and Parag M. Kanade Computer Science & Engineering Dept University of South Florida, Tampa FL 33620 @csee.usf.edu Abstract

More information

Automatic Modularization of ANNs Using Adaptive Critic Method

Automatic Modularization of ANNs Using Adaptive Critic Method Automatic Modularization of ANNs Using Adaptive Critic Method RUDOLF JAKŠA Kyushu Institute of Design 4-9-1 Shiobaru, Minami-ku, Fukuoka, 815-8540 JAPAN Abstract: - We propose automatic modularization

More information

MODELLING DOCUMENT CATEGORIES BY EVOLUTIONARY LEARNING OF TEXT CENTROIDS

MODELLING DOCUMENT CATEGORIES BY EVOLUTIONARY LEARNING OF TEXT CENTROIDS MODELLING DOCUMENT CATEGORIES BY EVOLUTIONARY LEARNING OF TEXT CENTROIDS J.I. Serrano M.D. Del Castillo Instituto de Automática Industrial CSIC. Ctra. Campo Real km.0 200. La Poveda. Arganda del Rey. 28500

More information

American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) ISSN (Print) , ISSN (Online)

American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) ISSN (Print) , ISSN (Online) American Scientific Research Journal for Engineering, Technology, and Sciences (ASRJETS) ISSN (Print) 2313-4410, ISSN (Online) 2313-4402 Global Society of Scientific Research and Researchers http://asrjetsjournal.org/

More information

Evolving SQL Queries for Data Mining

Evolving SQL Queries for Data Mining Evolving SQL Queries for Data Mining Majid Salim and Xin Yao School of Computer Science, The University of Birmingham Edgbaston, Birmingham B15 2TT, UK {msc30mms,x.yao}@cs.bham.ac.uk Abstract. This paper

More information

Data Mining Part 5. Prediction

Data Mining Part 5. Prediction Data Mining Part 5. Prediction 5.4. Spring 2010 Instructor: Dr. Masoud Yaghini Outline Using IF-THEN Rules for Classification Rule Extraction from a Decision Tree 1R Algorithm Sequential Covering Algorithms

More information

Fuzzy-Rough Feature Significance for Fuzzy Decision Trees

Fuzzy-Rough Feature Significance for Fuzzy Decision Trees Fuzzy-Rough Feature Significance for Fuzzy Decision Trees Richard Jensen and Qiang Shen Department of Computer Science, The University of Wales, Aberystwyth {rkj,qqs}@aber.ac.uk Abstract Crisp decision

More information

Look-Ahead Based Fuzzy Decision Tree Induction

Look-Ahead Based Fuzzy Decision Tree Induction IEEE TRANSACTIONS ON FUZZY SYSTEMS, VOL. 9, NO. 3, JUNE 2001 461 Look-Ahead Based Fuzzy Decision Tree Induction Ming Dong, Student Member, IEEE, and Ravi Kothari, Senior Member, IEEE Abstract Decision

More information

Using a fuzzy inference system for the map overlay problem

Using a fuzzy inference system for the map overlay problem Using a fuzzy inference system for the map overlay problem Abstract Dr. Verstraete Jörg 1 1 Systems esearch Institute, Polish Academy of Sciences ul. Newelska 6, Warsaw, 01-447, Warsaw jorg.verstraete@ibspan.waw.pl

More information

LEARNING WEIGHTS OF FUZZY RULES BY USING GRAVITATIONAL SEARCH ALGORITHM

LEARNING WEIGHTS OF FUZZY RULES BY USING GRAVITATIONAL SEARCH ALGORITHM International Journal of Innovative Computing, Information and Control ICIC International c 2013 ISSN 1349-4198 Volume 9, Number 4, April 2013 pp. 1593 1601 LEARNING WEIGHTS OF FUZZY RULES BY USING GRAVITATIONAL

More information

GRANULAR COMPUTING AND EVOLUTIONARY FUZZY MODELLING FOR MECHANICAL PROPERTIES OF ALLOY STEELS. G. Panoutsos and M. Mahfouf

GRANULAR COMPUTING AND EVOLUTIONARY FUZZY MODELLING FOR MECHANICAL PROPERTIES OF ALLOY STEELS. G. Panoutsos and M. Mahfouf GRANULAR COMPUTING AND EVOLUTIONARY FUZZY MODELLING FOR MECHANICAL PROPERTIES OF ALLOY STEELS G. Panoutsos and M. Mahfouf Institute for Microstructural and Mechanical Process Engineering: The University

More information

Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic

Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic Automatic Machinery Fault Detection and Diagnosis Using Fuzzy Logic Chris K. Mechefske Department of Mechanical and Materials Engineering The University of Western Ontario London, Ontario, Canada N6A5B9

More information

Web Data mining-a Research area in Web usage mining

Web Data mining-a Research area in Web usage mining IOSR Journal of Computer Engineering (IOSR-JCE) e-issn: 2278-0661, p- ISSN: 2278-8727Volume 13, Issue 1 (Jul. - Aug. 2013), PP 22-26 Web Data mining-a Research area in Web usage mining 1 V.S.Thiyagarajan,

More information

FUZZY INFERENCE SYSTEMS

FUZZY INFERENCE SYSTEMS CHAPTER-IV FUZZY INFERENCE SYSTEMS Fuzzy inference is the process of formulating the mapping from a given input to an output using fuzzy logic. The mapping then provides a basis from which decisions can

More information

Fuzzy Based Decision System for Gate Limiter of Hydro Power Plant

Fuzzy Based Decision System for Gate Limiter of Hydro Power Plant International Journal of Electronic and Electrical Engineering. ISSN 0974-2174 Volume 5, Number 2 (2012), pp. 157-166 International Research Publication House http://www.irphouse.com Fuzzy Based Decision

More information

Approximate Reasoning with Fuzzy Booleans

Approximate Reasoning with Fuzzy Booleans Approximate Reasoning with Fuzzy Booleans P.M. van den Broek Department of Computer Science, University of Twente,P.O.Box 217, 7500 AE Enschede, the Netherlands pimvdb@cs.utwente.nl J.A.R. Noppen Department

More information

Pedestrian Detection Using Correlated Lidar and Image Data EECS442 Final Project Fall 2016

Pedestrian Detection Using Correlated Lidar and Image Data EECS442 Final Project Fall 2016 edestrian Detection Using Correlated Lidar and Image Data EECS442 Final roject Fall 2016 Samuel Rohrer University of Michigan rohrer@umich.edu Ian Lin University of Michigan tiannis@umich.edu Abstract

More information

An Empirical Study of Hoeffding Racing for Model Selection in k-nearest Neighbor Classification

An Empirical Study of Hoeffding Racing for Model Selection in k-nearest Neighbor Classification An Empirical Study of Hoeffding Racing for Model Selection in k-nearest Neighbor Classification Flora Yu-Hui Yeh and Marcus Gallagher School of Information Technology and Electrical Engineering University

More information

An indirect tire identification method based on a two-layered fuzzy scheme

An indirect tire identification method based on a two-layered fuzzy scheme Journal of Intelligent & Fuzzy Systems 29 (2015) 2795 2800 DOI:10.3233/IFS-151984 IOS Press 2795 An indirect tire identification method based on a two-layered fuzzy scheme Dailin Zhang, Dengming Zhang,

More information

Irregular Interval Valued Fuzzy Graphs

Irregular Interval Valued Fuzzy Graphs nnals of Pure and pplied Mathematics Vol 3, No, 03, 56-66 ISSN: 79-087X (P), 79-0888(online) Published on 0 May 03 wwwresearchmathsciorg nnals of Irregular Interval Valued Fuzzy Graphs Madhumangal Pal

More information

A Predictive Controller for Object Tracking of a Mobile Robot

A Predictive Controller for Object Tracking of a Mobile Robot A Predictive Controller for Object Tracking of a Mobile Robot Xiaowei Zhou, Plamen Angelov and Chengwei Wang Intelligent Systems Research Laboratory Lancaster University, Lancaster, LA1 4WA, U. K. p.angelov@lancs.ac.uk

More information

Recent Progress on RAIL: Automating Clustering and Comparison of Different Road Classification Techniques on High Resolution Remotely Sensed Imagery

Recent Progress on RAIL: Automating Clustering and Comparison of Different Road Classification Techniques on High Resolution Remotely Sensed Imagery Recent Progress on RAIL: Automating Clustering and Comparison of Different Road Classification Techniques on High Resolution Remotely Sensed Imagery Annie Chen ANNIEC@CSE.UNSW.EDU.AU Gary Donovan GARYD@CSE.UNSW.EDU.AU

More information

COSC 6339 Big Data Analytics. Fuzzy Clustering. Some slides based on a lecture by Prof. Shishir Shah. Edgar Gabriel Spring 2017.

COSC 6339 Big Data Analytics. Fuzzy Clustering. Some slides based on a lecture by Prof. Shishir Shah. Edgar Gabriel Spring 2017. COSC 6339 Big Data Analytics Fuzzy Clustering Some slides based on a lecture by Prof. Shishir Shah Edgar Gabriel Spring 217 Clustering Clustering is a technique for finding similarity groups in data, called

More information

FUZZY INFERENCE. Siti Zaiton Mohd Hashim, PhD

FUZZY INFERENCE. Siti Zaiton Mohd Hashim, PhD FUZZY INFERENCE Siti Zaiton Mohd Hashim, PhD Fuzzy Inference Introduction Mamdani-style inference Sugeno-style inference Building a fuzzy expert system 9/29/20 2 Introduction Fuzzy inference is the process

More information

Fuzzy Modeling using Vector Quantization with Supervised Learning

Fuzzy Modeling using Vector Quantization with Supervised Learning Fuzzy Modeling using Vector Quantization with Supervised Learning Hirofumi Miyajima, Noritaka Shigei, and Hiromi Miyajima Abstract It is known that learning methods of fuzzy modeling using vector quantization

More information

Outlines. Fuzzy Membership Function Design Using Information Theory Measures and Genetic Algorithms. Outlines

Outlines. Fuzzy Membership Function Design Using Information Theory Measures and Genetic Algorithms. Outlines Fuzzy Membership Function Design Using Information Theory Measures and Genetic Algorithms Outlines Introduction Problem Statement Proposed Approach Results Conclusion 2 Outlines Introduction Problem Statement

More information

Fuzzy-Rough Sets for Descriptive Dimensionality Reduction

Fuzzy-Rough Sets for Descriptive Dimensionality Reduction Fuzzy-Rough Sets for Descriptive Dimensionality Reduction Richard Jensen and Qiang Shen {richjens,qiangs}@dai.ed.ac.uk Centre for Intelligent Systems and their Applications Division of Informatics, The

More information

Fuzzy Inventory Model without Shortage Using Trapezoidal Fuzzy Number with Sensitivity Analysis

Fuzzy Inventory Model without Shortage Using Trapezoidal Fuzzy Number with Sensitivity Analysis IOSR Journal of Mathematics (IOSR-JM) ISSN: 78-578. Volume 4, Issue 3 (Nov. - Dec. 0), PP 3-37 Fuzzy Inventory Model without Shortage Using Trapezoidal Fuzzy Number with Sensitivity Analysis D. Dutta,

More information

A Parallel Evolutionary Algorithm for Discovery of Decision Rules

A Parallel Evolutionary Algorithm for Discovery of Decision Rules A Parallel Evolutionary Algorithm for Discovery of Decision Rules Wojciech Kwedlo Faculty of Computer Science Technical University of Bia lystok Wiejska 45a, 15-351 Bia lystok, Poland wkwedlo@ii.pb.bialystok.pl

More information

Takagi-Sugeno Fuzzy System Accuracy Improvement with A Two Stage Tuning

Takagi-Sugeno Fuzzy System Accuracy Improvement with A Two Stage Tuning International Journal of Computing and Digital Systems ISSN (2210-142X) Int. J. Com. Dig. Sys. 4, No.4 (Oct-2015) Takagi-Sugeno Fuzzy System Accuracy Improvement with A Two Stage Tuning Hassan M. Elragal

More information

Argha Roy* Dept. of CSE Netaji Subhash Engg. College West Bengal, India.

Argha Roy* Dept. of CSE Netaji Subhash Engg. College West Bengal, India. Volume 3, Issue 3, March 2013 ISSN: 2277 128X International Journal of Advanced Research in Computer Science and Software Engineering Research Paper Available online at: www.ijarcsse.com Training Artificial

More information

Evaluating the Replicability of Significance Tests for Comparing Learning Algorithms

Evaluating the Replicability of Significance Tests for Comparing Learning Algorithms Evaluating the Replicability of Significance Tests for Comparing Learning Algorithms Remco R. Bouckaert 1,2 and Eibe Frank 2 1 Xtal Mountain Information Technology 215 Three Oaks Drive, Dairy Flat, Auckland,

More information

CPS331 Lecture: Fuzzy Logic last revised October 11, Objectives: 1. To introduce fuzzy logic as a way of handling imprecise information

CPS331 Lecture: Fuzzy Logic last revised October 11, Objectives: 1. To introduce fuzzy logic as a way of handling imprecise information CPS331 Lecture: Fuzzy Logic last revised October 11, 2016 Objectives: 1. To introduce fuzzy logic as a way of handling imprecise information Materials: 1. Projectable of young membership function 2. Projectable

More information

Fuzzy Mod. Department of Electrical Engineering and Computer Science University of California, Berkeley, CA Generalized Neural Networks

Fuzzy Mod. Department of Electrical Engineering and Computer Science University of California, Berkeley, CA Generalized Neural Networks From: AAAI-91 Proceedings. Copyright 1991, AAAI (www.aaai.org). All rights reserved. Fuzzy Mod Department of Electrical Engineering and Computer Science University of California, Berkeley, CA 94 720 1

More information

Effects of Three-Objective Genetic Rule Selection on the Generalization Ability of Fuzzy Rule-Based Systems

Effects of Three-Objective Genetic Rule Selection on the Generalization Ability of Fuzzy Rule-Based Systems Effects of Three-Objective Genetic Rule Selection on the Generalization Ability of Fuzzy Rule-Based Systems Hisao Ishibuchi and Takashi Yamamoto Department of Industrial Engineering, Osaka Prefecture University,

More information

C-NBC: Neighborhood-Based Clustering with Constraints

C-NBC: Neighborhood-Based Clustering with Constraints C-NBC: Neighborhood-Based Clustering with Constraints Piotr Lasek Chair of Computer Science, University of Rzeszów ul. Prof. St. Pigonia 1, 35-310 Rzeszów, Poland lasek@ur.edu.pl Abstract. Clustering is

More information

CHAPTER 5 FUZZY LOGIC CONTROL

CHAPTER 5 FUZZY LOGIC CONTROL 64 CHAPTER 5 FUZZY LOGIC CONTROL 5.1 Introduction Fuzzy logic is a soft computing tool for embedding structured human knowledge into workable algorithms. The idea of fuzzy logic was introduced by Dr. Lofti

More information

Stochastic propositionalization of relational data using aggregates

Stochastic propositionalization of relational data using aggregates Stochastic propositionalization of relational data using aggregates Valentin Gjorgjioski and Sašo Dzeroski Jožef Stefan Institute Abstract. The fact that data is already stored in relational databases

More information

What is all the Fuzz about?

What is all the Fuzz about? What is all the Fuzz about? Fuzzy Systems: Introduction CPSC 533 Christian Jacob Dept. of Computer Science Dept. of Biochemistry & Molecular Biology University of Calgary Fuzzy Systems in Knowledge Engineering

More information

Final Exam. Controller, F. Expert Sys.., Solving F. Ineq.} {Hopefield, SVM, Comptetive Learning,

Final Exam. Controller, F. Expert Sys.., Solving F. Ineq.} {Hopefield, SVM, Comptetive Learning, Final Exam Question on your Fuzzy presentation {F. Controller, F. Expert Sys.., Solving F. Ineq.} Question on your Nets Presentations {Hopefield, SVM, Comptetive Learning, Winner- take all learning for

More information

Enhancing K-means Clustering Algorithm with Improved Initial Center

Enhancing K-means Clustering Algorithm with Improved Initial Center Enhancing K-means Clustering Algorithm with Improved Initial Center Madhu Yedla #1, Srinivasa Rao Pathakota #2, T M Srinivasa #3 # Department of Computer Science and Engineering, National Institute of

More information

An Analysis of the Rule Weights and Fuzzy Reasoning Methods for Linguistic Rule Based Classification Systems Applied to Problems with Highly Imbalanced Data Sets Alberto Fernández 1, Salvador García 1,

More information

A Learning Algorithm for Tuning Fuzzy Rules Based on the Gradient Descent Method

A Learning Algorithm for Tuning Fuzzy Rules Based on the Gradient Descent Method A Learning Algorithm for Tuning Fuzzy Rules Based on the Gradient Descent Method Yan Shi*, Masaharu Mizumoto*, Naoyoshi Yubazaki** and Masayuki Otani** *Division of Information and Computer Sciences Osaka

More information

Correlation Based Feature Selection with Irrelevant Feature Removal

Correlation Based Feature Selection with Irrelevant Feature Removal Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology IJCSMC, Vol. 3, Issue. 4, April 2014,

More information

Fuzzy Signature Neural Networks for Classification: Optimising the Structure

Fuzzy Signature Neural Networks for Classification: Optimising the Structure Fuzzy Signature Neural Networks for Classification: Optimising the Structure Tom Gedeon, Xuanying Zhu, Kun He, and Leana Copeland Research School of Computer Science, College of Engineering and Computer

More information

Union and intersection of Level-2 fuzzy regions

Union and intersection of Level-2 fuzzy regions Union and intersection of Level- fuzzy regions Verstraete Jörg Systems esearch Institute, Polish Academy of Sciences ul. Newelska 6; 0-447 Warszawa; Polska Email: jorg.verstraete@ibspan.waw.pl Department

More information

Machine Learning & Statistical Models

Machine Learning & Statistical Models Astroinformatics Machine Learning & Statistical Models Neural Networks Feed Forward Hybrid Decision Analysis Decision Trees Random Decision Forests Evolving Trees Minimum Spanning Trees Perceptron Multi

More information

THE genetic algorithm (GA) is a powerful random search

THE genetic algorithm (GA) is a powerful random search 464 IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, VOL. 51, NO. 2, APRIL 2004 On Interpretation of Graffiti Digits and Characters for ebooks: Neural-Fuzzy Network and Genetic Algorithm Approach K. F. Leung,

More information

Modeling of ambient O 3 : a comparative study

Modeling of ambient O 3 : a comparative study Modeling of ambient O 3 : a comparative study Biljana Mileva-Boshkoska Abstract Air pollution is one of the biggest environmental concerns. Besides passive monitoring, the recent trend is shifting towards

More information

ECM A Novel On-line, Evolving Clustering Method and Its Applications

ECM A Novel On-line, Evolving Clustering Method and Its Applications ECM A Novel On-line, Evolving Clustering Method and Its Applications Qun Song 1 and Nikola Kasabov 2 1, 2 Department of Information Science, University of Otago P.O Box 56, Dunedin, New Zealand (E-mail:

More information