L01:
Only 20 minutes long, though I presented it quite badly as was rushed to lecture theatre. Need to rethink Bayesian section - it's not really making the points I want to make.


L02:
15 minutes short - need to add mean shift in. More images at end and some Bayesian would also help. What was there went well, though classification/regression need a figure.



L03:
10 minutes short. Abrupt jumping between topics. Has mistakes in that need fixing!
Should mention batching and circular arrays. Needs a demo and some images.
Exponential falloff needs a more direct explanation.



L04:
20 minutes short. Needs more figures to explain things, a more recent reading reference. Have time to add a recent paper.



L05:
5 minutes short. Example could be better presented, some more examples, mainly on importance sampling, but closest one to fine thus far.



L06:
20 minutes short, but failed to get demo ready in time. Should maybe revert to original plan and add Hamiltonian MC back. Lecture was a bit jumpy, bit think that was more tiredness than structure.



L07:
15 minutes short, though actually felt very full. Needs a bit of polish but is probably reasonable. Could add a little more - another example maybe?



L08:
Timing good, but not convinced they got it - too abstract at start. Need to rethink structure. Was admittedly really ill when giving it.



L15:
Only three students:-( Ran out of time, but somehow ended up about the right length. Bit around KL was a bit disconnected from rest - needs a stronger story. Missing examples, and LIGO not really a good one. Need to rethink intro - showing all the maths doesn't really help.



L16:
Only one student. I need a kitten. Lecture not great, but had good discussion at least! Missing demos, extras, clarity, images - everything that matters. Too much maths as well.



L17:
Reasonable length, 4 students turned up! need more images and some demos.


L19:
6 students, good length, got laughs. Could add back in the causality thing and it's good.
