Plans for the week of April 6-10
Dear all and welcome back. We hope you all had a great break and have been able to properly recharge your batteries.
We start now with generative methods for full and these will be keep us busy for the rest of the semester. The plan this week is (after a reminder from what where we ended before the break) is to focus on
Generative methods: energy models and Boltzmann machines. This entails
Restricted Boltzmann machines
Markov Chain Monte Carlo and Metropolis sampling
Gibbs sampling
Discussions of various Boltzmann machines
Implementation of simple restricted Boltzmann machines, own codes and using Pytorch
Reading recommendation: Goodfellow et al chapters 18.1-18.2, 20.1-20-7; To create Boltzmann machine using Keras, see Babcock and Bali chapter 4, see?https://github.com/PacktPublishing/Hands-On-Generative-AI-with-Python-and-TensorFlow-2/blob/master/Chapter_4/models/rbm.py
See also Foster, chapter 7 on energy-based models at?https://github.com/davidADSP/Generative_Deep_Learning_2nd_Edition/tree/main/notebooks/07_ebm/01_ebm
After our discussion of Boltzmann machines (this week and parts of next week) we will move on to variational autoencoders, and thereafter diffusion models, normalizing flows and finally GANS.
The tentative schedule for the rest of the semester is thus:
## April 13-17: Deep generative models
Reminder from previous week on Energy-based models and Langevin sampling
Variational Autoencoders
## April 20-24: Deep generative models
Variational autoencoders
## April 27 - May 1: ? Deep generative models
Diffusion models
## May 4-8: Deep generative models
Diffusion models and normalizing flows
## May 11-15:? No lecture May 14
Thursday May 14 is a public holiday in Norway, no lecture
## May 18-22:? GANS and summary of course
GANS
Summary?
Discussion and work on final project.
The jupyter-notebook for this week is at https://github.com/CompPhysics/AdvancedMachineLearning/blob/main/doc/pub/week11/ipynb/week11.ipynb.?
Best wishes to you all,
Morten, Oda and Ruben