Some consultation about the visualization results.

Hello, I am trying to use Bayes-opt to optimize 5-D problems, and all parameters are ‘range’ parameters. The generation strategy is using the default strategy, SOBOL for initial sampling, and GPEI for later optimization. The following are the generation strategy’s codes:

gs =  GenerationStrategy(
    steps=[
        # 1. Initialization step (does not require pre-existing data and is well-suited for 
        # initial sampling of the search space)
        GenerationStep(
            model=Models.SOBOL,
            num_trials=10,  # How many trials should be produced from this generation step
            min_trials_observed=3, # How many trials need to be completed to move to next model
            max_parallelism=5,  # Max parallelism for this step
            model_kwargs={"seed": 999},  # Any kwargs you want passed into the model
            model_gen_kwargs={},  # Any kwargs you want passed to `modelbridge.gen`
        ),
        # 2. Bayesian optimization step (requires data obtained from previous phase and learns
        # from all data available at the time of each new candidate generation call)
        GenerationStep(
            model=Models.GPEI,
            num_trials=-1,  # No limitation on how many trials should be produced from this step
            max_parallelism=3,  # Parallelism limit for this step, often lower than for Sobol
            # More on parallelism vs. required samples in BayesOpt:
            # https://ax.dev/docs/bayesopt.html#tradeoff-between-parallelism-and-total-number-of-trials
        ),
    ]
)

And 300 optimization loops are used, the following are the visualization results: My questions are whether the CV result proved our learned Gaussian model performed very poorly, except for doing more loops, how can I improve its performance? In addition, the contour plots look a little strange, would you like to tell me what is wrong?

[EugenHotaj]

The sampling bias might be due to the parameter constraints possibly not being set appropriately. The value of (lower, upper) = (-0.05, 1.9) while (ll, uu) = (-1.9, 0.05), so a huge part of the param space violates the constraints.

[EugenHotaj]

Closing this out as it seems to have been resolved. Please reopen if necessary.