Understanding f_classif in sklearn
Recently, to get better features, I used SelectKBest (score_func =, k = 20). f_classif computes anova f-value between function for classification problems. I've used it and got the best results. I found out that anova f-test calculates the relationship "between class variance" to "within class variance". But I have the following questions:
1) Does f_classif use a combination of functions to give an f-score ?. 2) Can I get pseudocode on how the SelectKBest function works? 3) How does f_classif work in sklearn ?. Thank you in advance.
+3
source to share