[ENH] Implement optimisation to `ElasticEnsemble` grid search

[MatthewMiddlehurst]

Describe the feature or idea you want to propose

From the code comments:

            # once the best parameter option has been estimated on the
            # training data, perform a final pass with this parameter option
            # to get the individual predictions with cross_cal_predict (
            # Note: optimisation potentially possible here if a GridSearchCV
            # was used previously. TO-DO: determine how to extract
            # predictions for the best param option from GridSearchCV)
            else:
                best_model = KNeighborsTimeSeriesClassifier(
                    n_neighbors=1,
                    distance=this_measure,
                    distance_params=grid.best_params_["distance_params"],
                    n_jobs=self._n_jobs,
                )
                preds = cross_val_predict(
                    best_model, full_train_to_use, y, cv=LeaveOneOut()
                )
                acc = accuracy_score(y, preds)

Describe your proposed solution

Explore if the grid search and/or random parameter search can be optimised to avoid doing an extra cross-validation when weighting.

Describe alternatives you've considered, if relevant

No response

Additional context

This can be done in a few ways possible, dont know what is correct:

• Redo CV if self.proportion_train_in_param_finding is less than 1
• Use the subsamples train set to generate accuracy estimate and always avoid an extra CV