Paper [1] is the most cited paper in Google Scholar. It is a survey paper that worth careful reading. Mainstream methods such as variable ranking, … are discussed in the paper. A check list is also summarized in the paper that is important for understanding feature selection.

Variance threshold is a widely used method provided in sklearn. The basic idea is removing features with low variance.

Pearson correlation coefficient is introduced in [2].

Mean decrease accuracy .

Tree-based feature selection …


[1] Guyon, I., & Elisseeff, A. (2003). An introduction to variable and feature selection. Journal of machine learning research, 3(Mar), 1157-1182.

[2] Hall, M. A. (1999). Correlation-based feature selection for machine learning.