Multivariate Correlation Analysis for Supervised Feature Selection in High-Dimensional Data

Abstract

The main theme of this dissertation focuses on multivariate correlation analysis on different data types and we identify and define various research gaps in the same. For the defined research gaps we develop novel techniques that address relevance of features to the target and redundancy of features amidst themselves. Our techniques aim at handling homogeneous data, i.e., only continuous or categorical features, mixed data, i.e., continuous and categorical features, and time series