We'll revisit statistical tests in preparation from the Neyman Pearson lemma https://drive.google.com/file/d/1P4jxjmvpBBXTr-vBDWtBFEeBwRerUMQ9/view?usp=sharing

We'll develop a confidence band for the empirical distribution - https://drive.google.com/file/d/1xXWyQmgI3oUhw-rlWZn7CcJa43elOTbX/view?usp=sharing

Whiteboard from today's meeting on Bayesian ML:

Cox:  P(A, B) = P(A|B)P(B) = P(B|A)P(A)
=>
P(A|B) = [P(B|A)P(A)] / P(B) (Bayes)
The ML model needs to be a generative model - statistically model that generate data
Example - a mixture of normal distribution.   Two normal distributions.  You choose which one you use in probability 1/2 and then you use to generate data.   
(n+1)! = (n+1)(n)!

https://en.wikipedia.org/wiki/Conjugate_prior

We revised Bayesian thinking and it's relation to ML.    https://drive.google.com/file/d/1CHxUPqtkyNWCklT7sgOC6I7YzAj29R-s/view?usp=sharing

This week we'll cover the theoretical basis for validation - https://drive.google.com/file/d/19TE0ZncmYP5KLnFPaWzffWFxp6Zxvkjp/view?usp=sharing

GD can be  regarded as a solution of an optimization problem with regularization  https://drive.google.com/file/d/1-5cdIOWzMW7LNWhBTHundK02GLtc2jtj/view?usp=sharing

This week we will discuss the expressive power of neural networks
https://drive.google.com/file/d/1hUmpgFX1YW0yOSl3iXh2L_f7n1eRLrjS/view?usp=sharing.