Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

feedback from piloting the concept/tutorials #69

Open
wavyparticle opened this issue Sep 20, 2024 · 0 comments
Open

feedback from piloting the concept/tutorials #69

wavyparticle opened this issue Sep 20, 2024 · 0 comments

Comments

@wavyparticle
Copy link

For the concepts, in general I think most of the stuff has been quite well written and it's really nice to have links to other resources

Single- vs. Multi-Objective Optimization

  • in the 2nd paragraph, I think it might be better to explain what is meant by "optimal"; I understood it as the input values that bring you closer to the goal but at first when I read it it wasn't clear to me

Frequentist Vs. Fully Bayesian Gaussian Process Models

  • in addition to the figure explaining the effect of length scale , it would be nice to also have a figure on the effect of output scale

  • would be nice to have an explanation of what "maximum a posteriori" means in addition to what it does in practice

  • under "Fully Bayesian GP Fitting", "tries of" in the 1st paragraph should be "tries to", and "the details of which" should be preceded by a comma (sry just a small grammar thing)

Single vs. Batch Optimization

  • for the 2nd figure, it helps me understand the earlier statement "expected value of the portion of the distribution outside of the best observed value relative to the optimization objective", but I think it might be a bit confusing to call the "best observed value relative to the optimization objective" y_max; also, for the black distribution, is it meant to be the distribution of points vertically below x_new, given by the light blue region? if so maybe it might be nice to change the colour of the curve to light blue as well; additionally, I think it would also be nice to add some annotation to the figure

Multitask Bayesian Optimization

  • not much here, most of it was quite clear

For the tutorials, I was only able to go through the 1st one (which was pretty straighforward) before I ran into the problem above; once that gets fixed I can go ahead and try the rest and if there's anything that comes to mind I can update my feedback here

extra: on the homepage under "if you're new to Bayesian optimization" maybe some links to videos by 3blue1brown would also be useful?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

1 participant