Bill Behrman
Bill Behrman
In some places $f$ is used to indicate the true function and in other places a variable in $p(f | \mathcal{D}_n)$. Some of the suggestions below are to try to...
[KR21a] M. Khan and H. Rue. "The Bayesian Learning Rule". In: (2022). arXiv: [2107.04562](https://arxiv.org/abs/2107.04562). page 265, line 30: in this section -> In this section page 266, line 28: $\exp(R(\boldsymbol{\theta}))$...
Book 1, formula 8.149: $\mathbf{Ł}(\boldsymbol{\theta}, q_n | \mathbf{y}_n)$ Book 1, formula 8.150: $\mathbf{Ł}(\boldsymbol{\theta}, q_n)$ Book 2, formulas 6.134 and 6.135: $\mathbf{Ł}(\boldsymbol{\theta}, q_n | \mathbf{y}_n)$ In my opinion, Book 2 is...
line 8: cals -> calls line 23: Student T -> Student line 29: discussed -> discuss
line 28: Did you mean "Student likelihood"? You use "Student distribution" below. line 38: as we discussed -> as we discuss
Better to have one of the following: Since the entropy terms $\mathbb{H}(q_n)$ are constant wrt $\boldsymbol{\theta}$, we can drop The entropy terms $\mathbb{H}(q_n)$ are constant wrt $\boldsymbol{\theta}$, so we can...
On pages 257 and 615 you use: kronecker product The rest of the book uses: Kronecker product (See also #458)
to find a matrix a negative definite matrix -> to find a negative definite matrix
If formula 8.108 is considered standard form, you may wish to use the same format for formula 8.112. Also, I think you could eliminate possible confusion by changing: $\mathbf{A}_{eq}\boldsymbol{\theta} =...