Model Creation & Testing #26

Use findings from ticket #4
Implement simple versions of each model and test their performance

Considerations:

• Generate fake data based of contract research
• Follow model testing best practices (train vs test, k-folds, ect...)
• Model evaluation and diagnostics (ex. goodness-of-fit, error metrics, residual diagnostics, predictor diagnostics, validation)
• Simpler the better
• Consistent results (think accuracy vs. precision)

Acceptance criteria:

• Python script to train and test each model
• Compare model performance and understand differences
• Be ready to present findings at next meeting (09/30)