A Hybrid Gene Selection Techniques for Cancer Classification.

      Аннотация: Due to the high dimensionality of microarray gene expression data, feature selection is an important task which improves the performance of the classification in the area of cancer diagnosis. In this article, innovative methods are proposed for selecting highly informative gene subsets of gene expression data that effectively classify the cancer data into tumorous and non-tumorous. A novel technique to identify signature biomarkers is also discussed.

Ключевые слова: accuracy, Feature Selection, miRNA, survival analysis, Cancer classification