Abstract:The inductivelearning of fuzzy rule based classification systems affects from exponentialgrowth of the fuzzy rule search space when the number of patterns and/orvariables becomes high. This growth types the learning process more problematicand, in most cases, it indications to problems of scalability (in relationshipsof the time and memory consumed) and/or complexity (with respect to the numberof rules obtained and the number of variables included in each rule). In thiswork, we propose a fuzzy association rule-based classification technique forhigh-dimensional problems based on three stages to find an accurate and compactfuzzy rule based classifier with a low computational cost. This techniqueparameters the order of the associations in the association rule extraction andconsiders the use of subgroup discovery based on an Improved Weighted RelativeAccuracy portion to preselect the most interesting rules earlier a geneticpost-processing process for rule selection and parameter change. The resultsfound done twenty-six real-world datasets of altered sizes and with differentnumbers of variables establish the effectiveness of the proposed method.
Keywords:Data Mining,Associative Classification, Classification, Fuzzy Association Rules, GeneticAlgorithms, Genetic Fuzzy Rule Selection, High-Dimensional Problems.