Research on the Application of Pattern Selection Algorithm in Bioinformatic Data Bases on Mutual Information

Pattern selection is an important part in the research fields of data mining and pattern recognition, especially for the high-dimensional data in the Bioinformatics. In this paper, a new pattern selection algorithm was proposed to finish pattern selection bases on Mutual Information. Pattern subset evaluation index was researched to ensure the best feature subset was selected. The algorithm bases on the correlation of patterns and label, as well as the redundancy between the patterns. Fuzzy Pattern Subset Evaluation Index was researched to make sure which is the best subset for the pattern subset evaluation. To verify the effectiveness of the method, some experiments were finished with the data of gene expression data (Leiden University) and UCI datasets. The experimental results indicate that the algorithm achieved better results.