The data used in this project is obtained from freezecasting.net : http://www.freezecasting.net/downloads.html
The schema xml file was downloaded and loaded with phpmyAdmin in a mysql database.
After loading the schema in to MySQL and connectin to python. The features of interest are loaded into a pandas dataframe, following the sql queries from the notebook in http://www.freezecasting.net/downloads.html
The dataframe was filtered and saved as csv in data folder.
Training examples and results are displayed in train.ipynb.
Trained models are saved in temp_models folder.
Train pipeline configuration is defined at custom_models.py
Grid search params are defined in grid_search_params.py