WWS586A-Machine-Learning-Policy-Analysis

Materials from my Princeton University course WWS586A: Machine Learning for Policy Analysis

Topics covered

Policy analysis in the digital age
Programming in R, R Notebooks & Github
Data acquisition and cleaning: APIs & webscraping
Natural language processing: tokenization, text processing, document-term matrix, TF-IDF weighting.
Introduction to statistical learning theory: bias-variance tradeoff, training, testing, cross-validation.
Model selection: regularization, LASSO, Bayesian LASSO.
Supervised learning: kNN, Naive Bayes, Decision Trees, SVMs, Neural Networks.
Unsupervised learning: k-means clustering, latent dirichlet allocation (topic models).
Special topics: machine learning and causal inference: Bayesian structural time series.