Three objective genetics-based machine learning for linguisitc rule extraction

摘要: Abstract This paper shows how a small number of linguistically interpretable fuzzy rules can be extracted from numerical data for high-dimensional pattern classification problems. One difficulty in the handling problems by rule-based systems is exponential increase with input variables. Another deterioration comprehensibility when they involve many antecedent conditions. Our task to design comprehensible high ability. formulated as combinatorial optimization problem three objectives: maximize correctly classified training patterns, minimize rules, and total We show two genetic-algorithm-based approaches. rule selection where are selected large prespecified candidate rules. The other genetics-based machine learning sets evolved genetic operations. These approaches search non-dominated respect objectives.