# Wrapping (feature selection)

Suppose we have some learning problem where the input domain has a large number of features. Wrapping is the process of selecting features for a learning algorithm by using that algorithm to evaluate how it performs on different subsets of features.