I read about Wrapped Method Feature Selection, I get that it is to look at the features then test them against the predictive model that we need then find if it has an effect or not and then decide to include it .

for Model Based Feature Selection, based on what I have read, it is the same !! Test the feature against predictive model and if it has effect we include it to the features list.

Now I am confused, are both Model Based Feature Selection & Wrapped Method Feature Selection are the same thing? or did I miss something or misunderstood the concept of Model Based Feature Selection?



According to many feature selection literature, wrapper feature selection and embedded feature selection techniques calculate the importance of features based on their predictive power. However, wrapper techniques compare different combinations of the proposed features and return the most accurate one and detect the interactions between features. The embedded methods select the features by learning the model and therefore, the feature selection runs within the learning process simultaneously.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

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