optuna.terminator.MedianErrorEvaluator

class optuna.terminator.MedianErrorEvaluator(paired_improvement_evaluator, warm_up_trials=10, n_initial_trials=20, threshold_ratio=0.01)[source]

An error evaluator that returns the ratio to initial median.

This error evaluator is introduced as a heuristics in the following paper:

A stopping criterion for Bayesian optimization by the gap of expected minimum simple regrets

Parameters:

paired_improvement_evaluator (BaseImprovementEvaluator) – The improvement_evaluator instance which is set with this error_evaluator.
warm_up_trials (int) – A parameter specifies the number of initial trials to be discarded before the calculation of median. Default to 10. In optuna, the first 10 trials are often random sampling. The warm_up_trials can exclude them from the calculation.
n_initial_trials (int) – A parameter specifies the number of initial trials considered in the calculation of median after warm_up_trials. Default to 20.
threshold_ratio (float) – A parameter specifies the ratio between the threshold and initial median. Default to 0.01.

Note

Added in v4.0.0 as an experimental feature. The interface may change in newer versions without prior notice. See https://github.com/optuna/optuna/releases/tag/v4.0.0.

Methods

evaluate(trials, study_direction)