Classifier Construction

Topics

This week’s assignments will guide you through the following topics:

• How to build and begin evaluating an XGBoost classifier

Reading

Please read the following:

• Read Sections 3.2 and 3.3 of An optimized XGBoost based diagnostic system for effective prediction of heart disease
• Supplemental reading How to Develop Your First XGBoost Model in Python

Replication task

• Build an XGBoost classifer using the cleaned heart data (don’t worry about parameter tuning for now)
• Explore the outputs of the trained model
• Calculate evaluation metrics as the authors do. Take a look at model accuracy, sensitivity, etc.

Tasks

Complete the following tasks:

• Complete the reading
• Build a classifier and calculate several evaluation metrics

Weekly Questions

Answer the following questions

• Why do we need to split our data into testing and training sets?
• We don’t always put as much weight on certain metrics as others. When would we care more about sensitivity? When would we care more about precision?
• In your own words, describe what a ROC curve is and why it’s useful. What is AUC and what does it tell us?