Multiobjective data mining from solutions by evolutionary multiobjective optimization

One research direction in the field of evolutionary multiobjective optimization (EMO) is a post-analytical process of non-dominated solutions in order to analyze the relationship between design variables and objective functions for optimization problems. For this purpose, data mining techniques have been used in some studies. From a practical point of view, this process itself should be considered as a multiobjective optimization problem. In this paper, multiobjective genetic fuzzy rule selection is applied to the post-analytical process of solutions obtained by EMO algorithms. First, multiple regions of interest are specified in the objective space. Each region with a number of solutions is handled as a different class. A set of patterns is generated by the labeled solutions. Second, a number of fuzzy if-then rules are generated by classification rule mining. Finally, an EMO algorithm is applied to combinatorial optimization of fuzzy if-then rules in order to obtain a number of non-dominated fuzzy classifiers with respect to accuracy and complexity. Through computational experiments using two engineering problems, we show that we can obtain various classifiers with a variety of complexity-accuracy tradeoff.

[1]  Peter J. Fleming,et al.  Genetic Algorithms for Multiobjective Optimization: FormulationDiscussion and Generalization , 1993, ICGA.

[2]  Sanaz Mostaghim,et al.  Heatmap Visualization of Population Based Multi Objective Algorithms , 2007, EMO.

[3]  Kalyanmoy Deb,et al.  Automated Innovization for Simultaneous Discovery of Multiple Rules in Bi-objective Problems , 2011, EMO.

[4]  Hisao Ishibuchi,et al.  Evolutionary Multiobjective Design of Fuzzy Rule-Based Classifiers , 2008, Computational Intelligence: A Compendium.

[5]  Kalyanmoy Deb,et al.  A fast and elitist multiobjective genetic algorithm: NSGA-II , 2002, IEEE Trans. Evol. Comput..

[6]  R. Barandelaa,et al.  Strategies for learning in class imbalance problems , 2003, Pattern Recognit..

[7]  Shinya Watanabe,et al.  Development of a multi-granularity design support system using hierarchical clustering for non-dominated solutions , 2010, 2010 World Automation Congress.

[8]  Masaya Nakata,et al.  Design Strategy Generation for a Sounding Hybrid Rocket via Evolutionary Rule-Based Data Mining System , 2017 .

[9]  Akira Oyama,et al.  Multidisciplinary and Multi-objective Design Exploration Methodology for Conceptual Design of a Hybrid Rocket , 2011 .

[10]  Francisco Herrera,et al.  Multiobjective genetic fuzzy rule selection of single granularity-based fuzzy classification rules and its interaction with the lateral tuning of membership functions , 2011, Soft Comput..

[11]  R. K. Ursem.  Multi-objective Optimization using Evolutionary Algorithms , 2009 .

[12]  C. Fonseca,et al.  GENETIC ALGORITHMS FOR MULTI-OBJECTIVE OPTIMIZATION: FORMULATION, DISCUSSION, AND GENERALIZATION , 1993 .

[13]  Kalyanmoy Deb,et al.  Data mining methods for knowledge discovery in multi-objective optimization: Part A - Survey , 2017, Expert Syst. Appl..

[14]  Hisao Ishibuchi,et al.  Fuzzy data mining: effect of fuzzy discretization , 2001, Proceedings 2001 IEEE International Conference on Data Mining.

[15]  Aravind Srinivasan,et al.  Innovization: innovating design principles through optimization , 2006, GECCO.

[16]  Shigeru Obayashi,et al.  Data Mining for Aerodynamic Design Space , 2005, J. Aerosp. Comput. Inf. Commun..

[17]  Kalyanmoy Deb,et al.  Towards automating the discovery of certain innovative design principles through a clustering-based optimization technique , 2011 .

[18]  Kalyanmoy Deb,et al.  Unveiling innovative design principles by means of multiple conflicting objectives , 2003 .

[19]  Daisuke Sasaki,et al.  Visualization and Data Mining of Pareto Solutions Using Self-Organizing Map , 2003, EMO.

[20]  Shigeru Obayashi,et al.  A New Design Method based on Cooperative Data Mining from Multi-Objective Design Space , 2009 .