An efficient feature selection framework based on information theory for high dimensional data

TL;DR: This survey provides a comprehensive overview of state-of-art feature selection techniques, categorizing them into five domains: subspace learning, sparse representation, information theory, evolutionary computational algorithms, and reinforcement learning, highlighting their strengths and limitations.

TL;DR: In this paper , a new filter-based feature selection (FS) framework based on the best-worst multi-attribute decision-making method was proposed and compared to two control groups: (a) no FS and (b) randomized algorithm.

TL;DR: Wang et al. as discussed by the authors proposed an online early terminated online streaming feature selection (OSFS-ET) framework, which can terminate the streaming features early before the end of streaming features and guarantee a competing performance with the currently selected features.

TL;DR: In this article , a new Student-t Kernelized Fuzzy Rough Set (SKFRS) model is proposed, which uses fuzzy divergence to evaluate uncertain information in the data and explores a newly defined feature evaluation function on the biases of the dynamic relation between the relevance and indispensability of features in feature selection process.

TL;DR: A set of simple, yet safe and robust non-parametric tests for statistical comparisons of classifiers is recommended: the Wilcoxon signed ranks test for comparison of two classifiers and the Friedman test with the corresponding post-hoc tests for comparisons of more classifiers over multiple data sets.

TL;DR: In this article, the maximal statistical dependency criterion based on mutual information (mRMR) was proposed to select good features according to the maximal dependency condition. But the problem of feature selection is not solved by directly implementing mRMR.

TL;DR: This work derives an equivalent form, called minimal-redundancy-maximal-relevance criterion (mRMR), for first-order incremental feature selection, and presents a two-stage feature selection algorithm by combining mRMR and other more sophisticated feature selectors (e.g., wrappers).

TL;DR: In this paper, variable and feature selection have become the focus of much research in areas of application for which datasets with tens or hundreds of thousands of variables are available, such as t...