A hybrid data gravitation based classification algorithm applied to gene expression data

Gene expression data classification problem has caught many researchers' attention in recent years. The main challenge mentioned in such problem is to overcome the high dimensionality in the limited sample sizes. A novel classification algorithm, data gravitation based classification model (DGC), shows significant performance in many general classification problems. Also, there is an important character of DGC to deal with gene expression data classification problem, feature weighing procedure which measures the importance of a feature by applying weight. In this study, we design a hybrid classifier based on the basic DGC model namely ADGC for Gene expression data classification problem. We use ANOVA as a filter to quickly decrease irrelevant, redundant and nuisance genes before apply DGC and improved simplified swarm optimization algorithm to optimize the feature weight for DGC. To practically evaluate the performance of the proposed method, a total ten gene expression datasets are used to test the performance of the proposed method, and corresponding results are compared with up-to-date works. Experimental results present that ADGC is effective for gene expression data classification problems.