Feature Importance of categorical variables by converting them into dummy variables (One-hot-encoding) can skewed or hard to interpret results. Here I present a method to get around this problem using H2O.
Geek Repo:Geek Repo
Github PK Tool:Github PK Tool
