The Blue Pill
This is a binary classification model to predict credit card approvals.
Models evaluated - Two-Class Boosted Decision Tree model,
Two-Class Neural Network
Two-Class Logistic Regression
After loading up the credit card data, columns are selected.
Key, Ethnicity, Male(Gender), and Drivers License are excluded from the model.
Then the data is cleaned up. 5% of the data (37 entries) are missing data and are removed from the data set.
After setting categorical variables and label (predictor), the data set is split 70:30 into training and test data.
Then the 3 models are trained, scored and evaluated.
