Plans for the week of April 22-26

Dear all, this week we continue our discussion of variational autoencoders (VAEs), their mathematics and programming implementations. We will discuss again possible paths for project 2, see for example https://github.com/CompPhysics/AdvancedMachineLearning/blob/main/doc/Projects/2024/Project2/pdf/Project2.pdf

Note that we have moved the deadline to June 7.  

Two excellent articles summarize VAEs and diffusion models and our lecture this coming Tuesday follows these articles. They are

1. Kingma and Welling, An Introduction to Variational Autoencoders, see https://arxiv.org/abs/1906.02691.

2. Calvin Luo gives an excellent link between VAEs and diffusion models, see https://calvinyluo.com/2022/08/26/diffusion-tutorial.html

Else, the lecture notes are at for example  https://github.com/CompPhysics/AdvancedMachineLearning/blob/main/doc/pub/week14/ipynb/week14.ipynb

These notes will be update during Monday April 22 (not final version yet).

Concerning the second project, for those of you who have planned to write one project only, feel free to proceed with that.

Irrespective of that, we would like you consider styling the report as a scientific report.

The guidelines we have established at https://github.com/CompPhysics/AdvancedMachineLearning/tree/main/doc/Projects/EvaluationGrading could be useful in structuring your report. This reflects also the way we grade. We have also added a lecture set by Anne Ruimy (director of EDP journals, https://fr.linkedin.com/in/anneruimy) on how to write effective titles and abstracts. See https://github.com/CompPhysics/AdvancedMachineLearning/tree/main/doc/Projects/WritingAbstracts for these lectures. Finally, at https://github.com/CompPhysics/AdvancedMachineLearning/tree/main/doc/Projects/2023/ProjectExamples you can find different examples of previous reports. See also the literature suggestions below.

For those of you who plan to write a second project, we would like to propose that you focus on generative methods, in particular those we have discussed during the lectures. These are

1. Boltzmann machines

2. Variational autoencoders and GANs

3. Diffusion models

Else, see the project description for more details.

best wishes to you all,

Keran, Morten and Ruben