Implement hyperparameter optimization via Gaussian search from scikit-optimize #1292
Conversation
| @@ -3451,6 +3451,137 @@ def randomized_search(self, param_distributions, X, y=None, cv=3, n_iter=10, par | |||
| def _convert_to_asymmetric_representation(self): | |||
| self._object._convert_oblivious_to_asymmetric() | |||
|
|
|||
| def gaussian_search(self, param_distributions, X, y=None, cv=3, n_random_starts=10, | |||
There was a problem hiding this comment.
the interface should be the same as in random_search and grid_search except for parameter distribution
| partition_random_seed=None, n_jobs=1, const_params={}, to_minimize_objective=True, | ||
| refit=True, train_size=0.8, verbose=True, plot=False): | ||
| import skopt | ||
| if n_calls<= 0: |
There was a problem hiding this comment.
please follow the codestyle everywhere
| @@ -3451,6 +3451,137 @@ def randomized_search(self, param_distributions, X, y=None, cv=3, n_iter=10, par | |||
| def _convert_to_asymmetric_representation(self): | |||
| self._object._convert_oblivious_to_asymmetric() | |||
|
|
|||
| def gaussian_search(self, param_distributions, X, y=None, cv=3, n_random_starts=10, | |||
There was a problem hiding this comment.
docs are needed for this function, so that the user will understand the sense of all the parameters
|
|
||
| self.model.set_params(**self.const_params) | ||
| self.model.set_params(**params_dict) | ||
| if self.search_by_train_test_split: |
There was a problem hiding this comment.
Train-test split must be done once for the whole parameter tuning process. Quantization must be done once or if quantization parameters are among the optimized ones, then it must be done every time when quantization changes.
| self.model.set_params(**params_dict) | ||
| if self.search_by_train_test_split: | ||
| if isinstance(self.X, Pool): | ||
| self.X = self.X.get_features() |
There was a problem hiding this comment.
This will not work in case if you have quantized pool, if there are categorical features or if there are texts.
| fold_count=self.cv, | ||
| partition_random_seed=self.partition_random_seed) | ||
| result = list(result["test-" + self.loss_function + "-mean"])[-1] | ||
| if not self.to_minimize_objective: |
There was a problem hiding this comment.
This value should not be specifier by the user, this is always fully specified by the evaluation metric.
|
|
||
| from skopt.utils import use_named_args | ||
| from skopt import gp_minimize | ||
| from skopt.space import Real, Categorical, Integer |
There was a problem hiding this comment.
please remove all unused imports
| @@ -3451,6 +3451,137 @@ def randomized_search(self, param_distributions, X, y=None, cv=3, n_iter=10, par | |||
| def _convert_to_asymmetric_representation(self): | |||
| self._object._convert_oblivious_to_asymmetric() | |||
|
|
|||
| def gaussian_search(self, param_distributions, X, y=None, cv=3, n_random_starts=10, | |||
There was a problem hiding this comment.
since the method requires skopt params, let's name it skopt_parameter_search
| random_state=None, n_calls=100, search_by_train_test_split=True, | ||
| partition_random_seed=None, n_jobs=1, const_params={}, to_minimize_objective=True, | ||
| refit=True, train_size=0.8, verbose=True, plot=False): | ||
| import skopt |
There was a problem hiding this comment.
please import only the things that are used inside the method
| params=params_dict, | ||
| fold_count=self.cv, | ||
| partition_random_seed=self.partition_random_seed) | ||
| result = list(result["test-" + self.loss_function + "-mean"])[-1] |
There was a problem hiding this comment.
this will work not for all loss_functions, plus loss_function might be None in this code
I hereby agree to the terms of the CLA available at: https://yandex.ru/legal/cla/?lang=en
This pull request is a part of a task of the ML Engineering course in the YDSA.