EvanClaes

Hello @saitcakmak, Thank you so much for you reply. If qNEHVI (and then I assume EHVI, qEHVI, NEHVI as well) can produce infeasible candidates, how is this typically dealt with...

@saitcakmak @Balandat, thanks. Really appreciate your input. Will try the deterministic approach first. Have a good day!

So I tried max's proposal. The code is complaining that I didn't provide 'batch_initial_conditions', although I did. Maybe in the wrong format? `# optimize candidates, _ = optimize_acqf( acq_function=acq_func, bounds=standard_bounds,...

Also tried Siat's suggestion, in which case I get the following: `models = [] for i in range(train_y.shape[-1]): models.append( SingleTaskGP(train_x, train_y[..., i : i + 1], outcome_transform=Standardize(m=1)) ) models.append(GenericDeterministicModel(evaluateCons1)) model...

Hello @Balandat, again thanks so much for your reply. Really love how much you guys are committing to (leading?) the open source community around BO. For the problems I'm faced...

@Balandat thanks for this. Will try it out. I'm curious if it would also be possible to do the sequential greedy optimization outside of `optimize_acqf`? If I execute it for...

Hello Eytan, Thanks for your feedback, it helped. The noise specification was indeed the problem. I was giving the model the noise of the observed values, not of the standardized...