epfml/ML_course

Typos in lecture 6a (GLM)

MaximeKjaer opened this issue · 0 comments

Hi,

I think I've spotted a few typos in the lecture notes for lecture 6a on Generalized Mixture Models:

  • Page 9, the probability of Gaussian distribution needs to be conditioned on both parameters image, not just on mu image. Alternatively, maybe a bold mu could be used to indicate the vector of usual parameters?
  • Page 14, μ and σ² are not the natural parameters, they're the usual parameters. η is the natural parameter.
  • Page 15, the log is missing a closing parenthesis, which should be right after h(yn)