We will be reading Pattern Recognition And Machine Learning by Chris Bishop, and the plan is to meet every 2 weeks and cover one chapter of the book each week. Because the chapters are pretty long it will likely work best if everyone reads the chapter beforehand, and then we use the session to discuss topics we didn't fully understand, or thought were particularly interesting. A different person will lead the session each week, be in charge of doing a few things:
- Giving a short (~10 mins) summary of the key content in the chapter at the start of the session, and maybe highlighting a few key equations and plots.
- Moderating the discussion and making sure everyone has a chance to speak and ask questions.
- Deciding on the exercises for the group for that week, listing them in the exercises.md file a few days before the session, and hopefully have the solutions to hand for the session (solutions for many of the exercises in the book can be found here) and also here.
Below is the provisional schedule for the sessions and leaders (we may want to skip a few of the later chapters):
| Section | Leader | Time |
|---|---|---|
| 5.1-5.3 (Neural Networks) | Magnus | 11/02/21 16:00 |
| 5.4-5.7 (Neural Networks) | Mauricio | 25/02/21 |
| 6.1-6.3 (Kernel Methods) | Tom | 11/03/21 |
| TBC | TBC | TBC |