Combining Support Vector Machines and the t-statistic for Gene Selection in DNA Microarray Data Analysis
This paper proposes a new gene selection (or feature selection) method for DNA microarray data analysis. In the method, the t-statistic and support vector machines are combined efficiently. The resulting gene selection method uses both the data intrinsic information and learning algorithm performance to measure the relevance of a gene in a DNA microarray. We explain why and how the proposed method works well. The experimental results on two benchmarking microarray data sets show that the proposed method is competitive with previous methods. The proposed method can also be used for other feature selection problems.
KeywordsSupport Vector Machine Gene Selection Linear Support Vector Machine Ranking Criterion Recursive Feature Elimination
Unable to display preview. Download preview PDF.