Thanks! And, for sure! There is a really great tool for hyper tuning some really black box models (Neural Networks) that track progress. I can definitely make a video on that topic. Stay tuned! :) It probably be in python.
Ahh yes. The dreaded rank deficiency error. There can be a multitude of reasons why that appears 1) don’t have enough data 2) quality of data is poor 3) need to scale model differently (center data and scale) 4) one of your variables is collinear with another (remove this variable) I’m sure there are some other EDA analysis to do on your data, but those are the steps I’d check out first.
