Picture the scenario
I apply fastai tabular to a Kaggle dataset to get a baseline
Results in very poor performance
This leaves me with two questions
I would love it if someone illustrated further or described how they do it for improved performance?
Why ?
What can I do to improve performance