Feature Selection in Data Mining - Approaches Based on Information Theory 12+

📒 In many predictive modeling tasks, one has a fixed set of observations fromwhich a vast, or even infinite, set of potentially predictive features can becomputed. Of these features, often only a small number are expected to beuseful in a predictive model. Models which use the entire set of features willalmost certainly overfit on future data sets.The book presents streamwise feature selection which interleaves the processof generating new features with that of feature testing. Streamwise featureselection scales well to large feature sets. The book also describes how to usestreamwise feature seleciton in multivariate regressions.It includes a review of traditional feature selecitions in a general frameworkbased on information theory, and compares these methods with streamwisefeature selection on various real and synthetic data sets. This book isintended to be used by researchers in machine learning, data mining, andknowledge discovery.