作者: Huiwen Zeng , H. J. Trussell
关键词:
摘要: Feature selection is the process of selecting effective subsets features that are in performing a given task. We propose an approach using penalty function combined with neural network to select subset from collection while maintaining performance possible larger set. The related mixed-norm has proven successful pruning networks. new shown work on test cases known redundancy and be feature for practical problems.