0
$\begingroup$

I am using a binary logistic regression (a type of probabilistic statistical classification model, is used to predict a likelihood of belonging to a class (True, False)). I have 4 features and I want to know what is the best subset of using these features to reach the best result? should I try all of possible subset of these features? Is there any scientific solution for this purpose? I don't have sufficient knowledge about this. Any help will be highly appreciated.

$\endgroup$

marked as duplicate by Xi'an, Nick Cox, gung - Reinstate Monica, kjetil b halvorsen, chl Feb 2 '15 at 16:34

This question has been asked before and already has an answer. If those answers do not fully address your question, please ask a new question.

0
$\begingroup$

Since your number of features is small, you can try all possible combinations.

In general, regularization is often used for feature selection:

http://openclassroom.stanford.edu/MainFolder/DocumentPage.php?course=MachineLearning&doc=exercises/ex5/ex5.html

$\endgroup$

Not the answer you're looking for? Browse other questions tagged or ask your own question.