作者: JingTao Yao , Ming Zhang
DOI: 10.1007/11548669_22
关键词:
摘要: We present a study on rough set based approach for feature selection. Instead of using significance or support, Parameterized Average Support Heuristic (PASH) considers the overall quality potential rules. It will produce rules with balanced support distribution over all decision classes. Adjustable parameters PASH can help users different levels approximation needs to extract predictive that may be ignored by other methods. This paper finetunes heuristic and provides experimental results PASH.