Informative variable identifier: Expanding interpretability in feature selection

• Published in: Pattern recognition 2020-02, Vol.98, p.107077, Article 107077
• Main Authors: Muñoz-Romero, Sergio, Gorostiaga, Arantza, Soguero-Ruiz, Cristina, Mora-Jiménez, Inmaculada, Rojo-Álvarez, José Luis
• Format: Article
• Language: English
• Subjects: Classification; Explainable machine learning; Feature selection; Interpretability; Resampling
• ISSN: 0031-3203; 1873-5142
• DOI: 10.1016/j.patcog.2019.107077

•Interpretability of the solution is provided by a novel feature selection algorithm.•Relevant, redundant and non-informative input variables are identified.•Analysis of weights learned by resampling allows to clarify relations among variables.•Improvement in the interpretability of the results and in classification performance. There is nowadays an increasing interest in discovering relationships among input variables (also called features) from data to provide better interpretability, which yield more confidence in the solution and provide novel insights about the nature of the problem at hand. We propose a novel feature selection method, called Informative Variable Identifier (IVI), capable of identifying the informative variables and their relationships. It transforms the input-variable space distribution into a coefficient-feature space using existing linear classifiers or a more efficient weight generator that we also propose, Covariance Multiplication Estimator (CME). Informative features and their relationships are determined analyzing the joint distribution of these coefficients with resampling techniques. IVI and CME select the informative variables and then pass them on to any linear or nonlinear classifier. Experiments show that the proposed approach can outperform state-of-art algorithms in terms of feature identification capabilities, and even in classification performance when subsequent classifiers are used.