Formulas in the exam

We have got several questions regarding the exam and to which degree you have remember formulas, in particular from the part on supervised learning. I have made a document trying to answer

- formulas you have to know, and formulas you don't need to remember

- how to interpret some formulas

- relationships between various learning algorithms

I have placed the document in the practical session, here.

Enjoy reading, Jan Tore

Updated, May 23: I corrected one typo. The loss in multinomial logistic regression is "general cross-entropy" not "binary cross-entropy".