I think there might be a typo in equation 12, where the log of the posterior is calculated.
In equation 6 there is a minus before E_D and it looks like it has been overlooked, when using the log prior in equation 12.
When trying to derivate the log posterior by using Bayes Theorem I almost got the same solution as in this article but was not able to find my error with the sign.
While reading Pattern Recognition and Machine Learning from Bishop(2006) I saw on page 153 equation 3.55 I found support of my theory that there might be a missing sign.
I created a PR #12. If I'm wrong please just close the request and the PR.
krasserm
commented
4 years ago
Greetings,
I think there might be a typo in equation 12, where the log of the posterior is calculated.
In equation 6 there is a minus before E_D and it looks like it has been overlooked, when using the log prior in equation 12.
When trying to derivate the log posterior by using Bayes Theorem I almost got the same solution as in this article but was not able to find my error with the sign.
While reading Pattern Recognition and Machine Learning from Bishop(2006) I saw on page 153 equation 3.55 I found support of my theory that there might be a missing sign.
I created a PR #12. If I'm wrong please just close the request and the PR.