Skip to content

Latest commit

 

History

History
114 lines (88 loc) · 6.19 KB

probabilistic-ml.md

File metadata and controls

114 lines (88 loc) · 6.19 KB

Advanced Probabilistic Machine Learning

The SDML book club will start discussing Probabilistic Machine Learning: Advanced Topics by Kevin Murphy (https://probml.github.io/pml-book/book2.html) in January 2023. Please read the format and logistics document for information about the structure of the series.

Notes and videos

Below are links to the meetups, recordings, and any additional materials.
All of the videos will be posted to Ryan's YouTube channel: https://www.youtube.com/c/ITConnected/videos

Schedule

Session 1: Kickoff and Chapter 1, Introduction
The kickoff meeting will be Tuesday, January 31, 2023, from 6:00-7:30 pm Pacific
We will meet other people in small breakout rooms and discuss chapter 1 and sections 2.1-2.2 of PML book 2.
Here are the notes for sections 2.1-2.2.
People also shared a couple of their favorite resources for Bayes' Theorem:

video of session 1 content

Session 2: Chapter 2, Probability
The meetup will be Tuesday, February 7, 2023, 6:00-7:30 pm PST.
Discussion leader: Neal Fultz
We will finish the discussion of probability.
Author's references include:

video of session 2

Session 3: Chapter 3, Statistics
The session will be Tuesday, February 14, 2023, 6:00-7:30 pm PST.
Discussion leader: Roger Stager
We will start the discussion of statistics in chapter 3.
video

Session 4: Chapter 3, Statistics
The event will be Tuesday, February 21, 2023, 6:00-7:30 pm PST.
Discussion leader: Roger Stager
We will continue the discussion of statistics, and we expect to need another session to finish chapter 3.
video

Session 5: Chapter 3, Statistics
The meetup will be Tuesday, February 28, 2023, 6:00-7:30 pm PST.
Discussion leader: Jayanth Raman
We finish going through statistics with sections 3.6-3.11.
video

Session 6: Chapter 4, Graphical models
The session will be Tuesday, March 7, 2023, 6:00-7:30 pm PST.
Discussion leader: Dev Vidhani
We will start the discussion of chapter 4 about graphical models.

Session 7: Chapter 4, Graphical models
The event will be Tuesday, March 14, 2023, 6:00-7:30 pm PDT.
Discussion leader: Dev Vidhani
We plan to finish chapter 4 about graphical models.

Session 8: Chapter 5, Information theory
The meetup will be Tuesday, March 21, 2023, 6:00-7:30 pm PDT.
Discussion leader: Christian Zuniga
We plan to start and discuss most of the information theory chapter.

Session 9: Chapter 6, Optimization
The session will be Tuesday, March 28, 2023, 6:00-7:30 pm PDT.
Discussion leader: Phawis Thammasorn
Christian will finish sections 5.5 and 5.6, then Phawis will begin the discussion of optimization.

Session 10: Chapter 6, Optimization
The meeting will be Tuesday, April 4, 2023, 6:00-7:30 pm PDT.
Discussion leader: Dev Vidhani
We will briefly look at preconditioned optimizers (including Adam) in section 8.4.6 of Kevin Murphy's intro book, then continue our discussion of optimization from section 6.5.

Session 11: Chapter 6, Optimization
The meetup will be Tuesday, April 11, 2023, 6:00-7:30 pm PDT.
Discussion leader: Dev Vidhani
We will finish discussing optimization.

Part II -- Inference

Session 12: Chapter 7, Inference algorithms: an overview
The event will be Tuesday, April 18, 2023, 6:00-7:30 pm PDT.
Discussion leader: Ted Kyi
We start the second part of the book with an overview of Bayesian Inference.

Session 13: Chapter 8, Gaussian filtering and smoothing
The session will be Tuesday, May 2, 2023, 6:00-7:30 pm PDT.
Discussion leader: Mya Bakhova
We cover the Kalman filter and its extensions as we talk about Gaussian filtering and smoothing on state space models, which are represented by a simple linear chain of hidden states over time.

Session 14: Chapter 9, Message passing algorithms
The meetup will be Tuesday, May 9, 2023, 6:00-7:30 pm PDT.
Discussion leader: Christian Zuniga
We will discuss Hidden Markov Models, belief propagation, and other topics related to message passing algorithms, which allow posterior inference for general probabilistic graphical models.

Remainder of the initial schedule (dates TBD):

  • Chapter 10, Variational inference
  • Chapter 11, Monte Carlo methods
  • Chapter 12, Markov chain Monte Carlo
  • Chapter 13, Sequential Monte Carlo