Some more minor issues about udlbook HOT 2 CLOSED

Thanks so much. You are of course technically right about the denominator of Bayes' rule (eq. 9.11) but I think the way that I have written it is the more common usage. I think on balance changing the variable has to \phi' probably has just as much scope for confusion, so I left it.

The last missing reference is a forward reference to the normalizing flows chapter. That's not public yet. Hoping to get a few more chapters out in the next few weeks.

Thank you again!

from udlbook.

"MNIST-1D" is used everywhere except Chapter 10 and the table of contents, where "MNIST1D" is used.
In pages 163 and 165 there are missing references to a figure, probably fig.8.1.
p.163: "two, and “valid” padding The third hidden layer " (Missing full stop).
p.170: "input contributes to three of the outputs When we" (Missing full stop).
p.174: "fine-tuned for object detection" (Missing full stop)
p.178: "and LeCun et al. (1989a) Early applications" (Missing full stop)
In page 263 another reference to a figure is missing, "must be a contraction mapping (see figure ??)."

from udlbook.