template<Numeric NumericType, Derived< CosanModel > Model, Derived< CosanMetric< NumericType >> Metric, Derived< Splitter > Split>
class Cosan::GridSearch< NumericType, Model, Metric, Split >
Hyperparameter tuning for supervised models that have one or more hyperparameter(s) to tune Input required: estimator: class Model&, a model whose hyperparameters need to be tuned; metric: class Metric&, a metric to use in cross-validation split: Split & splitter method;* paramGrid: a vector of hyperparameters combination. For each entry of the vector, it corresponds to one choice of hyperparameter combination; Ouput: call .GetBestParams() to get the best hyperparameters combination. the choice of the hyper-parameter in paramGrid that forms the most accurate model
Definition at line 26 of file gridsearch.h.