‘Travaux dirigés’ of Machine Learning

Teaching to students of first year of Master’s degree in Computer Science at École normale supérieure of Lyon. Exercises on foundations of machine learning, both theoretical, through mathematical proofs, and applied, using Python notebooks. —

Topics covered

• Lesson 1: Empirical vs true risk, Bayes classifier;
• Lesson 2: Conditional expectation, kNN, regresso-gram;
• Lesson 3: Linear Regression, Ordinary Least Squares, Singular Value Decomposition;
• Lesson 4: Logistic Regression, LDA, Gradient Descent, Newton’s method;
• Lesson 5: Clustering, k-Means(++), Hierarchical Clustering, Spectral Clustering;
• Lesson 6: Dimensionality reduction (PCA, Dictionary Learning, Autoencoders);
• Lesson 7: Trees, random forests, ensemble methods;
• Lesson 8: Density estimation, Gaussian Mixture Models;
• Lesson 9: Stochastic Gradient Descent;
• Lesson 10: DL Intro, Convolutional networks, Batch-Norm, Residual Layers;
• Lesson 11: Generative models, Flow Matching.