Journal
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
Volume 43, Issue 12, Pages 4396-4410Publisher
IEEE COMPUTER SOC
DOI: 10.1109/TPAMI.2020.3002843
Keywords
Feature extraction; Mutual information; Redundancy; Markov processes; Computational complexity; Correlation; Laplace equations; Feature selection; filter methods; markov chains
Funding
- Engineering and Physical Sciences Research Council (EPSRC) [EP/N035305/1]
- Italian Ministry of Education, Universities and Research (MIUR) Dipartimenti di Eccellenza 2018-2022
- EPSRC [EP/N035305/1] Funding Source: UKRI
Ask authors/readers for more resources
The study introduces a filtering feature selection framework that considers feature subsets as paths in a graph to address relevance and redundancy principles. Two different interpretations are used to evaluate the values of feature subsets, and a simple unsupervised strategy is proposed to cut the ranking and provide the subset of features to keep. The experiments show that the Infinite Feature Selection framework performs better in almost any situation compared to widely-known comparative approaches.
We propose a filtering feature selection framework that considers subsets of features as paths in a graph, where a node is a feature and an edge indicates pairwise (customizable) relations among features, dealing with relevance and redundancy principles. By two different interpretations (exploiting properties of power series of matrices and relying on Markov chains fundamentals) we can evaluate the values of paths (i.e., feature subsets) of arbitrary lengths, eventually go to infinite, from which we dub our framework Infinite Feature Selection (Inf-FS). Going to infinite allows to constrain the computational complexity of the selection process, and to rank the features in an elegant way, that is, considering the value of any path (subset) containing a particular feature. We also propose a simple unsupervised strategy to cut the ranking, so providing the subset of features to keep. In the experiments, we analyze diverse settings with heterogeneous features, for a total of 11 benchmarks, comparing against 18 widely-known comparative approaches. The results show that Inf-FS behaves better in almost any situation, that is, when the number of features to keep are fixed a priori, or when the decision of the subset cardinality is part of the process.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available