search

UBC-MDS / fifa-potential

Supervised machine learning model to predict potential rating of players in FIFA 22
https://ubc-mds.github.io/fifa-potential/high-potential-fifa-prediction-report.html
Other
0 stars 0 forks source link

EDA file added #7

Closed WaleedMahmood1 closed 7 months ago

WaleedMahmood1 commented 7 months ago

@meretelutz I have a question regarding the visualizations. I was wondering, would it be a better idea to create the visualizations on the non-scaled features rather than the scaled versions of the features? Or maybe I might have missed something and we should be creating the visualizations on the scaled features😅

meretelutz commented 7 months ago

@WaleedMahmood1 I had the same question myself! Our original visualizations used the raw data. But then I was looking at the project Tiff gave us as an example and saw that she was using the scaled data for visualizations. So I thought I would do the same. I'm not entirely sure why we would chose one over the other, but once the data is scaled there's quite a bit of overlap, so maybe I'll go back and play around with it again.

srfrew commented 7 months ago

To chime in here I would vote for unscaled data so we have a sense of what everything is without modification, but don't have much rationale besides that.

meretelutz commented 7 months ago

I agree, I think that makes more sense. I changed it above.