Penalized Logistic Regression in Gene Expression Analysis

A typical task in gene expression analysis is the classification of biological samples into two alternative categories. A statistical procedure is needed which, based on the expression profiles measured, allows to compute the probability that a new sample belongs to a certain class. The feature of high-dimensionality and small sample sizes makes this statistical task very challenging. Standard logistic regression fails in most instances because of condition problems. A state-of-the-art overview on penalized logistic regression approaches, including the choice of penalty functions and regularization parameters, is given.