Understanding a multivariate semi-metric in the search strategies for attributes subset selection
Gustavo Sosa Cabrera, Miguel García-Torres, Santiago Gómez Guerrero, Christian E. Schaerer, Federico Divina · Proceeding Series of the Brazilian Society of Computational and Applied Mathematics · 2018
In classification tasks, a feature (i.e. independent variable) is considered relevant, irrelevant or redundant according to the information contained about the class (i.e. dependent variable). Feature selection consists of finding the minimal set of relevant features such that the classification error is optimized.[...]