Feature selection for high dimensional data: An evolutionary filter approach.

摘要: Problem statement: Feature selection is a task of crucial importance for the application of machine learning in various domains. In addition, recent increase data dimensionality poses a severe challenge to many existing feature approaches with respect efficiency and effectiveness. As an example, genetic algorithm effective search that lends itself directly selection; however this direct hindered by data dimensionality. Therefore adapting cope high data becomes increasingly appealing. Approach: study, we proposed adapted version genetic algorithm can be applied dimensional data. The approach is based essentially on variable length representation scheme and set modified proposed genetic operators. To assess effectiveness approach, it cues phrase selection compared its performance number ranking which are always applied task. Results Conclusion: results provide experimental evidences the effectiveness