From explanations to feature selection: assessing SHAP values as feature selection mechanism

  • Wilson E. Marcílio-Jr Unesp
  • Danilo Eler Unesp

Resumo


Explainability has become one of the most discussed topics in machine learning research in recent years, and although a lot of methodologies that try to provide explanation to black-box models or try to be more transparent regarding internal decisions have been proposed to address such issue, little discussion has been made on the pre-processing steps involving the pipeline of development of machine learning solutions, such as feature selection. In this work, we evaluate a game theoretic approach used to explain output of any machine learning model, SHAP, as a feature selection mechanism. In the experiments, we show that besides being able to explain decisions of a model, it achieves better results than three commonly used feature selection algorithms.
Publicado
07/11/2020
MARCÍLIO-JR, Wilson E.; ELER, Danilo. From explanations to feature selection: assessing SHAP values as feature selection mechanism. In: CONFERENCE ON GRAPHICS, PATTERNS AND IMAGES (SIBGRAPI), 33. , 2020, Evento Online. Anais [...]. Porto Alegre: Sociedade Brasileira de Computação, 2020 . p. 303-310.