1 paper
Kayhan Behdin, Riade Benbaki, Peter Radchenko +1
We study the high-dimensional linear regression problem with categorical predictors that have many levels. We propose a new estimation approach, which performs model compression vi…