Best model if you have numerical, categorical, Text and Image features to predict same target variable


(Nisar Ahamed) #1

Fairly new to the ML and DL universe. Was wondering what would be the best approach to create a model that should a predict a target variable based on features that have numerical, categorical, Text and Image elements.

One example for such a problem would be the latest Kaggle Competition.


(Jeremy Howard) #2

Any DL model should work fine for that. We cover multi-modal learning in http://course.fast.ai so maybe start there?


(Nisar Ahamed) #3

Thanks a lot. I believe that is covered in the Part 2 of Cutting Edge DL Course?. Currently I am about to finish Part 1. Will Check it out for sure. Thanks again.


(Jeremy Howard) #4

Yes part 2 would be a good idea for learning more about this. Feel free to ask in that forum if you finish that course and still aren’t sure.