- Using the data from your project topic, choose a categorical target
- If necessary, change continiuous variable to categorical via a threshold, a la "data['affair'] = data['affairs'] > 1
- Choose features to use to predict the target
- Prepare an exploratory analysis of the features and target, include graphs.
- Run a logistic regression on your target using the features
- Make a prediction based on an interesting input set
- Save your model
- Prepare a test/train split and score your model against the training set
- Turn in the complete analysis as an IPython Notebook