Lecture calendar

Zoom lecture room: link

Notes will sometimes be posted in advance, but all notes are subject to change.

                  LecturesReading and Materials
Week 1

Lecture 1: Welcome to ECE324, Intro to observational fairness in ML (up to slide 17). Zoom recording

Lecture 2: Intro to observational fairness in ML, cont'd. Zoom recording

Reading: Barocas et al, Ch. 2; Corbett-Davies and Goel, The Measure and Mismeasure of Fairness: A Critical Review of Fair Machine Learning

Reading: CIML Ch. 9.1-9.2

Week 2

Lecture 1: Quick review of Maximum Likelihood, Bayesian inference, torch_coin_manual.py. Zoom recording

Lecture 2: Bayesian inference (and unicorns). Code: coin_bayes.py, Intro to Generative Models. Zoom recording

Lecture 3: Causal inference. Zoom recording.

Reading: CIML Ch. 9.1-9.2. PyTorch Tutorial another PyTorch Tutorial

Video: Bayesian inference about unicorns

Just for fun: You Can Load a Die, But You Can't Bias a Coin

Reading: Shalizi Ch. 18-19

Week 3

Lecture 1: Fairness and causality. Zoom recording

Lecture 2: Causality and ML. Zoom recording

Lecture 3: Word embeddings. Zoom recording

Reading: Kusner et al., Counterfactual Fairness, NeuriPS 2017

Reading: Mitchell at al., Model Cards for Model Reporting, FAT* 19

Reading: Shoelkopf et al., Toward Causal Representation Learning, Proc. of the IEEE, 2021

Video: Yoshua Bengio, Towards Causal Representation Learning

Week 4

Lecture 1: Word embeddings cont'd. Transformers I. Zoom recording

Lecture 2: Transformers II. Zoom recording

Lecture 3: minGPT. OpenAI API playground. Zoom recording

Reading: Attention is All You Need, NeuRIPS 2017

Reading: The Illustrated Transformer

Just for fun: GPT-3 Creative fiction

Week 5

Lecture 1: Transformers III. Zoom recording

Lecture 2:

Reading

Week 6

Lecture 1: GANs. Zoom recording

Lecture 2: Variational Autoencoders. Zoom recording

Lecture 3: Summary of training VAE and GAN, VAE implementation, GAN implementation. Wasserstein GAN. Zoom recording.

Reading:

(Very optional) proof of the Kantorovich-Rubinstein Theorem: pdf and video by Dmitry Panchenko

Reading: Weng, From Autoencoder to Beta-VAE, blog post, 2018

Reading: Weng, From GAN to WGAN, blog post, 2017

Week 7

Lecture 1: Wasserstein GAN, Zoom recording

Lecture 2: Graph embeddings and Graph Neural Networks I. Zoom recording (Quercus recording also available). Better quality Zoom recording

Lecture 3: Q&A

Keep readings: Weng, From GAN to WGAN, blog post, 2017

Week 8

Lecture 2: Graph embeddings and Graph Neural Networks I cont'd, Graph embeddings and Graph Neural Networks II. Zoom recording

Lecture 3: Graph embeddings and Graph Neural Networks II, cont'd. Recommender Systems with GNN. Zoom recording

Reading:

Week 9

Lecture 1: Generalization in deep networks and adversarial examples. Zoom recording.

Lecture 2: Generalization in deep networks and adversarial examples II, how ConvNets see. Zoom recording.

Lecture 3: How ConvNets See cont'd. Zoom recording.

Reading:

Week 10

Lecture 1: Reinforcement Learning with Policy Gradients. Zoom recording

Lecture 2: AI Ethics. Zoom recording

Lecture 3: Inductive bias and metalearning. No Free Lunch Theorem recording, Zoom recording

Reading:

Reading:

Reading: