Introduction to Machine Learning

University of California, Berkeley, Fall 2024

Welcome to CS 189/289A! This class covers theoretical foundations, algorithms, methodologies, and applications for machine learning. Topics may include supervised methods for regression and classication (linear models, trees, neural networks, ensemble methods, instance-based methods); generative and discriminative probabilistic models; deep learning models including CNNs, transformers, graph neural networks for vision and language tasks; and Markovian models for reinforcement learning and robotics.

Textbook: Deep Learning by Bishop and Bishop.

Professors will post slides prior to lecture at this Google Drive folder (for faster access). The material here is redundant with the website, but it may take up to a day or two for the website to get updated with the slides after lecture. The “Post-Lecture” subfolder contains updates to slides that the professors may make right after lecture. In addition, lecture recordings will be uploaded automatically to this Haas Panopto folder.

If you have not been added to EdStem, you may join through this link. The Gradescope code is DKPWZW.

Note: The topics for future lectures, discussions, and HWs are tentative and may be moved around, changed, or removed.

Overview

Week 1

8/29 Lecture 1 Introduction and Logistics (Slides) Reading 1.1-1.2.4
  Homework 1 Math Review (Due 9/11 9/12 11:59pm) PDF / Files / LaTeX Template

Week 2

9/3 Lecture 2 Maximum Likelihood Estimation (Slides) Reading
2-2.1.2 (rules of probability: sum, product)
2.1.6 (independent RVs)
2.2-2.2.1 (probability densities in continuous spaces)
2.3-2.3.2 (univariate Gaussian, likelihood)
3-3.1.3 (Bernoulli, binomial, multinomial, MLE)
  Discussion 0 Math Pre-Requisites Review Worksheet / Solutions
9/5 Lecture 3 Multivariate Gaussians (Slides) Reading 3-3.2.3 (multivariate Gaussians: geometry, moments, covariance forms)

Week 3

9/10 Lecture 4 Linear Regression 1 (Slides) Reading 4-4.1.4
  Discussion 1 MLE & Gaussians Worksheet / Solutions
9/12 Lecture 5 Linear Regression 2 (Slides) Reading
1.25-1.26 (regularization, model selection)
2.1.3 (Bayes theorem)
2.6 (Bayesian modeling)
4.1.6 (regularization for linear regression)
9.2.2 (Lasso regularization)
  Homework 2 Linear/Logistic Regression & Classification (Due 9/25 11:59pm) PDF / Files / LaTeX Template

Week 4

9/17 Lecture 6 Classification - Generative & Discriminative (Slides) Reading 5.1, 5.2.1, 5.2.2, 5.2.4, 5.3
  Discussion 2 Regularization & MAP Worksheet / Solutions
9/19 Lecture 7 Logistic Regression & Neural Networks (Slides) Reading 5.3.1, 5.4.1-5.4.4, 6.1-6.2

Week 5

9/24 Lecture 8 Backpropagation and Gradient Descent 1 (Slides) Reading 6.2, 6.4, 7.1, 7.2, 5.4.4
  Discussion 3 Classification & Logistic Regression Worksheet / Solutions
9/26 Lecture 9 Backpropagation and Gradient Descent 2 (Slides) Reading 5.4.4, 8
  Homework 3 Neural Networks (Due 10/9 11:59pm) PDF / Files / LaTeX Template

Week 6

10/1 Lecture 10 Neural Networks - CNNs, Batch Norm, & ResNets (Slides) Reading 7.4, 9.5, 10
  Discussion 4 Neural Networks and Backpropagation Worksheet / Solutions
10/3 Lecture 11 Neural Networks - Attention & Transformers (Slides) Reading 5.3, 12.1

Week 7

10/8 Lecture 12 Dimensionality Reduction & PCA (Slides) Reading 16.1
  Discussion 5 Convolution and Attention Worksheet / Solutions
10/10 Lecture 13 t-SNE (Slides)  
  Homework 4 Dimensionality Reduction & Decision Theory (Due 10/25 11:59pm) PDF / Files

Week 8

10/15 Lecture 14 Clustering Reading 15.1 (not including 15.1.1), 15.2
  Discussion 6 Dimensionality Reduction Techniques  
10/16 Midterm (7-9pm, Location TBD)  
10/17 Lecture 15 Multiway Classification, Decision Theory, & Model Evaluation  

Week 9

10/22 Lecture 16 Nearest Neighbors & Metric Learning  
  Discussion 7 Clustering and Decision Theory  
10/24 Lecture 17 Decision Trees & Ensembling  
  Homework 5 Bias/Variance, Nearest Neighbors, Decision Trees (Due 11/6 11:59pm)  

Week 10

10/29 Lecture 18 Bias-Variance Tradeoff & Over/Under-Fitting  
  Discussion 8 Bias/Variance and Nearest Neighbors  
10/31 Lecture 19 Hidden Markov Models & Graphical Models 1  

Week 11

11/5 Lecture 20 Hidden Markov Models & Graphical Models 2  
  Discussion 9 Decision Trees and HMMs Intro  
11/7 Lecture 21 Markov Decision Processes  
  Homework 6 Markovian Models & Reinforcement Learning (Due 11/20 11:59pm)  

Week 12

11/12 Lecture 22 Reinforcement Learning  
  Discussion 10 HMMs Advanced and MDPs  
11/14 Lecture 23 Robotics and Machine Learning  

Week 13

11/19 Lecture 24 Graph Neural Networks & Rotational Equivariance 1  
  Discussion 11 MDPs & Reinforcement Learning  
11/21 Lecture 25 Graph Neural Networks & Rotational Equivariance 2  
  Homework 7 Graph Neural Networks & Applications of Deep Learning (Due 12/4 11:59pm)  

Week 14

11/26 Lecture 26 Language and Vision Applications  
11/28 No Lecture (Thanksgiving)  

Week 15

12/3 Lecture 27 Special Topics - Causality  
  Discussion 12 Graph Neural Networks  
12/5 Lecture 28 Special Topics - Computational Biology  

Week 16 (RRR Week)

No lectures or discussions this week

Week 17 (Finals Week)

12/17 Final Exam (8-11am, location TBD)