Remember error (in shared image) related to ‘categorical variables’ ? We used
train_cats in lecture 2 to convert strings to pandas categories.
This CatBoost can also handle categorical variables directly and gives great results. I haven’t studied the underlying algorithm yet. But looks worth exploring.
Here is the link – https://www.analyticsvidhya.com/blog/2017/08/catboost-automated-categorical-data/
@groverpr , is this working well for you?
I am trying. Previously I was working on local machine. Seems like I need aws to run this.
Here is someone who tried and achieved good results with CatBoost – https://www.kaggle.com/aharless/simple-catboost-cv-lb-281
@groverpr Cool, thanks. Why do you need AWS to use it?
Parameter tuning takes time on my local machine.