Course Material
This page hosts all lecture slides and supplementary reading materials for each topic covered in the course.
Topic 1: Introduction
Click to view slides in browser
Supplementary Material
Topic 2: PGM Representation
Click to view slides in browser
Supplementary Material
- Probabilistic Graphical Models (Textbook [4]), Chapters 3 & 4
- CS228 (Probabilistic Graphical Models) Notes
Topic 3: PGM Inference
Click to view slides in browser
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 7.4
- Probabilistic Graphical Models (Textbook [4]), Chapter 9
- CS228 (Probabilistic Graphical Models) Notes
Topic 4: PGM Learning
Click to view slides in browser
Supplementary Material
- Probabilistic Graphical Models (Textbook [4]), Chapters 16 & 17
- A Note on the EM Algorithm
- CS228 (Probabilistic Graphical Models) Notes
Topic 5: Autoregressive Models
Click to view slides in browser
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 22
- Harvard NLP Tutorial: The Annotated Transformer
- CSE599i: Neural Autoregressive Density Estimation (NADE)
- CSE599i: Transformers
- Self-Attention from Scratch (Sebastian Raschka)
- Introduction to Positional Encoding (Machine Learning Mastery)
Topic 6: Variational Autoencoder
Click to view slides in browser
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 21
- CSE599i: VAE Notes
- CS236 (Deep Generative Models): VAE Notes
Topic 7: GAN
Click to view slides in browser
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 26
- CSE599i: GAN Notes
- Original GAN Paper (Goodfellow et al.)
- WGAN (Wasserstein GAN) Paper
- WGAN-GP (Improved WGAN) Paper
Topic 8: Normalizing Flow
Click to view slides in browser
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 23
- CS236 (Deep Generative Models): Normalizing Flow Notes
- CSE599i: Normalizing Flow Notes
- Neural Ordinary Differential Equations (Neural ODE) Paper
Topic 9: Energy-Based Models
Download Slides (Part 1) Download Slides (Part 2)
Click to view slides (Part 1) in browser
Click to view slides (Part 2) in browser
Supplementary Material
Topic 10: Score-Based Models
Download Slides (Animated PPTX) Download Slides (Static PDF)
Click to view slides (PDF) in browser
Format Recommendation
The PPTX file is highly recommended as the slides contain animations and GIFs. If you view the PDF, you will encounter some overlapping elements and meaningless figures due to these animations.
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 25
- Yang Song's Blog: Generative Modeling by Estimating Gradients of the Data Distribution
- CSE599i: Score-Based Models & Denoising Notes
Topic 11: Flow Matching
Click to view slides in browser
Supplementary Material
- To Be Added
Topic 12: Diffusion Models
Part 1: Introduction
Click to view slides in browser
Part 2: Advanced Concepts (Stanford)
Download Slides (Animated PPTX) Download Slides (Static PDF)
Click to view slides (PDF) in browser
Format Recommendation for Part 2
For Part 2, the PPTX file is highly recommended as the slides contain animations. If you view the PDF, you will encounter some overlapping elements and meaningless figures due to these animations.
Supplementary Material
- Probabilistic Machine Learning (Textbook [2]), Chapter 25
- Denoising Diffusion Probabilistic Models (DDPM) Paper
Topic 13: Evaluation of Generative Models
Click to view slides in browser
Supplementary Material
Topic 14: Parameter-Efficient Fine-Tuning (PEFT)
Click to view slides in browser
Supplementary Material
- To Be Added
Topic 15: Multi-Modal Models
Click to view slides in browser
Supplementary Material
- To Be Added