Computational intelligence methods and data understanding

Experts in machine learning and fuzzy system frequently identify understanding the data with the use of logical rules. Reasons for inadequacy of crisp and fuzzy rule-based explanations are presented. An approach based on analysis of probabilities of classification p(Ci|X;ρ) as a function of the size of the neighborhood ρ of the given case X is presented. Probabilities are evaluated using Monte Carlo sampling or – for some models – using analytical formulas. Coupled with topographically correct visualization of the data in this neighborhood this approach, applicable to any classifiers, gives in many cases a better evaluation of the new data than rule-based systems. Two real life examples of such interpretation are

[1]  Ryszard S. Michalski,et al.  A theory and methodology of inductive learning , 1993 .

[2]  Gregory Piatetsky-Shapiro,et al.  Knowledge discovery in databases: 10 years after , 2000, SKDD.

[3]  Aiko M. Hormann,et al.  Programs for Machine Learning. Part I , 1962, Inf. Control..

[4]  Sankar K. Pal,et al.  Neuro-Fuzzy Pattern Recognition , 1999 .

[5]  Ivan Bratko,et al.  Machine Learning and Data Mining; Methods and Applications , 1998 .

[6]  Wlodzislaw Duch,et al.  Optimization and Interpretation of Rule-based Classifiers , 2000, Intelligent Information Systems.

[7]  John Yen,et al.  Industrial Applications of Fuzzy Logic and Intelligent Systems , 1995 .

[8]  J. Kruskal.  Nonmetric multidimensional scaling: A numerical method , 1964 .

[9]  David J. Spiegelhalter,et al.  Machine Learning, Neural and Statistical Classification , 2009 .

[10]  Włodzisław Duch,et al.  NEURAL METHODS FOR ANALYSIS OF PSYCHOMETRIC DATA , 2000 .

[11]  O. Mangasarian,et al.  Robust linear programming discrimination of two linearly inseparable sets , 1992 .

[12]  Detlef Nauck,et al.  Foundations Of Neuro-Fuzzy Systems , 1997 .

[13]  Antoine Naud.  INTERACTIVE DATA EXPLORATION USING MDS MAPPING , 2000 .

[14]  James J. Buckley,et al.  On the equivalence of neural nets and fuzzy expert systems , 1999 .

[15]  Christopher J. Merz,et al.  UCI Repository of Machine Learning Databases , 1996 .

[16]  Włodzisław Duch,et al.  Neural optimization of linguistic variables and membership functions , 1999, ICONIP'99. ANZIIS'99 & ANNES'99 & ACNN'99. 6th International Conference on Neural Information Processing. Proceedings (Cat. No.99EX378).

[17]  Włodzisław Duch,et al.  Methodology of extraction , optimization and application of logical rules , 1999 .

[18]  Visakan Kadirkamanathan,et al.  Statistical Control of RBF-like Networks for Classification , 1997, ICANN.

[19]  D. M. Titterington,et al.  Statistics and Neural Networks , 2000, Technometrics.

[20]  Wlodzislaw Duch,et al.  A new methodology of extraction, optimization and application of crisp and fuzzy logical rules , 2001, IEEE Trans. Neural Networks.

[21]  Shigeo Abe,et al.  Neural Networks and Fuzzy Systems , 1996, Springer US.

[22]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[23]  L. Breiman.  Bias-variance, regularization, instability and stabilization , 1998 .

[24]  Włodzisław Duch.  A framework for similarity-based classification methods , 2000 .

[25]  K. Gr,et al.  A general purpose separability criterion for classification systems , 1999 .

[26]  Catherine Blake,et al.  UCI Repository of machine learning databases , 1998 .

[27]  Simon Haykin,et al.  Neural Networks: A Comprehensive Foundation , 1998 .

[28]  Krzysztof Grabczewski,et al.  Extraction of logical rules from backpropagation networks , 1998 .

[29]  Norbert Jankowski,et al.  New developments in the Feature Space Mapping model , 2000 .

[30]  Wlodzislaw Duch,et al.  Feature space mapping as a universal adaptive system , 1995 .