Som Dhulipala comments

Results 26 comments of


                                            Som Dhulipala

Refactor GP to consider ADAM

@grmnptr Thanks for refactoring the GP classes! Looks great!

Refactor GP to consider ADAM

@zachmprince @grmnptr We can select a batch of the training data for each optimization iteration. I can implement this as well. Is there a way to randomly select m indices...

Refactor GP to consider ADAM

Everything remaining the same, ADAM and TAO are giving pretty comparable results. Blue is ADAM and orange is TAO. Left is the standard deviation and right is the mean.

Refactor GP to consider ADAM

> > @zachmprince @grmnptr We can select a batch of the training data for each optimization iteration. I can implement this as well. Is there a way to randomly select...

Refactor GP to consider ADAM

@grmnptr Thanks for your comments. I have addressed all of them except one.

Refactor GP to consider ADAM

@grmnptr Something in your `GaussianProcessHandler.md` file is failing. I didn't change anything here.

Parallel active learning in Monte Carlo simulations with Gaussian Process

Please do not review this yet. I need to make a few more additions/changes. (wondering whether I can add labels...)

Parallel active learning in Monte Carlo simulations with Gaussian Process

> You cant add labels sorry. Which label were you thinking of adding? Maybe "do not review" or "do not merge"? But if I cannot remove them, I'd prefer not...

Parallel active learning in Monte Carlo simulations with Gaussian Process

> You would not be able to remove them but also we would do it for you and doubt it would become an issue > > You can also turn...

Parallel active learning in Monte Carlo simulations with Gaussian Process

@zachmprince @grmnptr This is ready for review. I am only fixing some minor doc issues.