In this challenge, the dataset contains 81 columns. So do we have to apply feature selection method to eliminate useless columns. Or dow we have to use all columns?
Doubt in housing price prediction challenge
hey @abhaygarg2001,
This is an experiment that you need to perform. Features depend upon the model you use.
All features might be useful for a model but there may be a case where only a couple of features can work much better and provide improved results.
Or you may need to create more features out of these features to get more better results.
So , just experiment as much as you can.
I am sure that doing a proper feature engineering can you provide you with state of the art results.
Thank You 
.