A Novel Granularity Optimal Feature Selection based on Multi-Variant Clustering for High Dimensional Data

Clustering is the most complex in multi/high dimensional data because of sub feature selection from overall features present categorical sources. Sub set be aggressive approach to decrease dimensionality mining data, identification patterns. Main aim behind with respect optimal and redundancy. In-order compute redundant/irrelevant high sample exploration based on calculation granular described this document. Propose aNovel Granular Feature Multi-variant Genetic Algorithm (NGFMCGA) model evaluate performance results implementation. This main consists two phases, first phase, theoretic graph grouping procedure divide into different clusters, second select strongly representative related each cluster matching subset features. Features concept are independent proposed clustering have probability processing increasing quality useful features.Optimal improves accuracy classification, describes better applied publicly sets it compared traditional supervised evolutionary approaches