smac.optimizer.configuration_chooser.turbo_chooser

Classes

TurBOChooser(scenario, stats, runhistory, ...)

Interface to train the EPM and generate next configurations with TurBO:
class smac.optimizer.configuration_chooser.turbo_chooser.TurBOChooser(scenario, stats, runhistory, runhistory2epm, model, acq_optimizer, acquisition_func, rng, restore_incumbent=None, random_configuration_chooser=<smac.optimizer.configuration_chooser.random_chooser.ChooserNoCoolDown object>, predict_x_best=False, min_samples_model=1, length_init=0.8, length_min=0.00390625, length_max=1.6, success_tol=3, failure_tol_min=4, n_init_x_params=2, n_candidate_max=5000)[source]

Bases: smac.optimizer.configuration_chooser.epm_chooser.EPMChooser

Interface to train the EPM and generate next configurations with TurBO:

D. Eriksson et al. Scalable Global Optimization via Local Bayesian Optimization https://papers.nips.cc/paper/2019/file/6c990b7aca7bc7058f5e98ea909e924b-Paper.pdf

Parameters

  • length_init (float) – Initialized length after restarting

  • length_min (float) – If the subspace length is smaller than length_min, TurBO will restart

  • length_max (float) – The maximum length of subspace

  • success_tol (int) – Number of successful suggestions (suggested points become incumbent) required for expanding subspace

  • failure_tol_min (int) – The minimum number of failure suggestions (suggested points fails to become incumbent) required for shrinking subspace

  • n_init_x_params (int) – how many configurations will be used at most in the initial design (X*D). Used for restarting the subspace

  • n_candidate_max (int) – Maximal Number of points used as candidates

choose_next(incumbent_value=None)[source]

Choose next candidate solution with TuRBO

Parameters

incumbent_value (float) – Cost value of incumbent configuration (required for acquisition function); If not given, it will be inferred from runhistory or predicted; if not given and runhistory is empty, it will raise a ValueError.

Return type

Iterator

previous

smac.optimizer.configuration_chooser.random_chooser

next

smac.optimizer.pSMAC