An efficient model-based method of hyperparameter tuning that builds a probabilistic model to choose the most promising configurations based on past results.
An efficient model-based method of hyperparameter tuning that builds a probabilistic model to choose the most promising configurations based on past results.