Weekly update for week 46

Dear All, 

a brief update on the plans for this week and project 3. First thx to everybody for proposing so many great variants of project 3. We really look forward to see your results.

Project 3 is now set up, see for example https://compphysics.github.io/MachineLearning/doc/Projects/2018/Project3/html/Project3-bs.html (or do a git pull or scroll down https://compphysics.github.io/MachineLearning/doc/web/course.html

 

It has three variants, one where you define the data, two where you study the credit card data and three to study the solution of the diffusion equation using neural networks. 

 

We have set the deadline to December 14 and that is because I am still hoping we can run the workshop. The week of November 26-30 is not possible for us due to too many collisions with exams.  

The reason I would like to propose to still run the workshop with ditto poster session, is simply because there are so many exciting alternatives which you folks have proposed. And I believe that you will be equally excited as me to see all these projects. 

So, I would like to propose either December 10 or 11, alternatively the 13th. The 12th is blocked.

Let me know if these days work, else we will have to give up.

 

Lectures this week: Finally we will see our ML methods from another perspective than the frequentist approach we have taken. Prof Nils Hjort from the Math dept, and a great expert on Bayesian statistics, will explain to us how the elegant formalism of Bayesian statistics can be used to derive probability distribution functions and thereby make predictions. 

Bayesian theory is a very elegant approach and has gained a lot of traction in the last years.

 

Next week we end our topics for this semester with unsupervised learning, mainly an overview and wrap up everything on Thursday the 29th. 

 

Best wishes to you all,

Morten

Publisert 14. nov. 2018 01:21 - Sist endret 14. nov. 2018 01:22