Gene Extraction for Cancer Diagnosis by Support Vector Machines An Improvement and Comparison with Nearest Shrunken Centroid Method

摘要

A cancer diagnosis by using the DNA microarray data faces many challenges the most serious one being the presence of thousands of genes and only several dozens (at the best) of patient's samples. Thus, making any kind of classification in high-dimensional spaces from a limited number of data is both an extremely difficult and a prone to an error procedure. The improved Recursive Feature Elimination with Support Vector Machines (RFE-SVMs) is introduced and used here for an elimination of less relevant genes and just for a reduction of the overall number of genes used in a medical diagnostic. The paper shows why and how the, usually neglected, penalty parameter C influence classification results and the gene selection of RFE-SVMs. With an appropriate parameter C chosen, the reduction in a diagnosis error is as high as 37% on the colon cancer data set. The results suggest that with a properly chosen parameter C, the extracted genes and the constructed classifier will ensure less over-fitting of the training data leading to an increase accuracy in selecting relevant genes.