bayesian-optimization
bayesian-optimization copied to clipboard
thuijskens
Python code for bayesian optimization using Gaussian processes
In line 48: expected_improvement[sigma == 0.0] == 0.0 It should be: expected_improvement[sigma == 0.0] = 0.0
In gp.py (when finding the next sample): ```next_sample = sample_next_hyperparameter(expected_improvement, model, yp, greater_is_better=True, bounds=bounds, n_restarts=100)``` Here `greater_is_better` is True means that its trying to maximize the loss function instead of...
Hey love the implementation it is very clean. I am having some issues making it work for integer-valued hyperparameters and am wondering if you have any suggestions?
Can you please specify under which license has the code been published? Thanks, Daniel
Metadata
Owner
Metadata
Python code for bayesian optimization using Gaussian processes