Become a member
Sign in
Rushi Sheth
Rushi Sheth

Rushi Sheth

22 Following
20 Followers
  • Profile
  • Claps
  • Highlights

Highlighted by Rushi Sheth

See more

From Handling categorical features: get_dummies, OneHotEncoder and multicollinearity by Cheng Ji

Secondly, you may not be able to drop one dummy variable in order to avoid multicollinearity after fitting OHE as there are no separate variables in the dataframe. One way to address this problem is to use regularization (Lasso) with the model to drop redundant dummies. As there are two dummies which provide the exact same information, one of them must be dropped when minimizing the regularization term in the cost function.

Claps from Rushi Sheth

See more

The Only One Skill That, If You Have, Will Completely Change Your Life

Oskar Nowik