Coevolutionary Fuzzy Attribute Order Reduction With Complete Attribute-Value Space Tree

Since big data sets are structurally complex, high-dimensional, and their attributes exhibit some redundant irrelevant information, the selection, evaluation, combination of those large-scale pose huge challenges to traditional methods. Fuzzy rough have emerged as a powerful vehicle deal with uncertain fuzzy in problems that involve very large number variables be analyzed short time. In order further overcome inefficiency algorithms data, this paper we present new coevolutionary attribute reduction algorithm (CFAOR) based on complete attribute-value space tree. A tree model decision table is designed adaptively prune optimize The similarity multimodality can extracted satisfy needs users better convergence speed classification performance. Then, rule generate series chains form an efficient cascade entropy threshold. Finally, performance CFAOR assessed set benchmark contain complex high dimensional datasets noise. experimental results demonstrate achieve higher average computational efficiency accuracy, compared state-of-the-art Furthermore, applied extract different tissues surfaces dynamical changing infant cerebral cortex it achieves satisfying consistency medical experts, which shows its potential significance for disorder prediction cerebrum.