I came back here in December 2023 to go through the data with formal ML and DS training. I was able to get the RMSE down to about 137 tickets when using a Random Forest Regressor.
Code that I used for my Senior Scholar's project at SJC. This is the code only, and will not run without the data. Since the data is proprietary, it has not been uploaded here. The purpose of this repo is just to demonstrate the methods that I used. This was before I had any formal training in ML, so it's not very well optimized.