• To learn the terminology associated with machine learning (e.g., train-test splits, cross-validation)
• To learn how to execute an end-to-end machine learning pipeline using the scikit-learn library
• To learn various techniques for inspecting how a machine learning model makes its decisions
IMPORTANT: This module focuses on “traditional” machine learning models like logistic regressors and random forest classifiers. It does not cover state-of-the-art deep learning like large language models (LLMs) or foundational models. However, the last lecture will show how the concepts covered in this module are prerequisite for these advanced model architectures.
• Crash Course on Machine Learning
• Classification with an Image Dataset: Skin Cancer
• Regression with a Time-Series Dataset: Parkinsonian Gait
• Refining Our Image Classifier
• Refining Our Time-Series Regressor
• Motivation
• Naïve Baselines
• Feature Importance
• SHAP Values
• Quantifying Bias
• Primer on Image CNNs