:py:mod:`desdeo_emo.surrogatemodels` ==================================== .. py:module:: desdeo_emo.surrogatemodels .. autoapi-nested-parse:: This module provides implementations of EAs which can be used for training surrogate models. Submodules ---------- .. toctree:: :titlesonly: :maxdepth: 1 BioGP/index.rst EvoDN2/index.rst EvoNN/index.rst Problem/index.rst Package Contents ---------------- Classes ~~~~~~~ .. autoapisummary:: desdeo_emo.surrogatemodels.BioGP desdeo_emo.surrogatemodels.EvoNN desdeo_emo.surrogatemodels.EvoDN2 desdeo_emo.surrogatemodels.surrogateProblem .. py:class:: BioGP(training_algorithm: Type[desdeo_emo.EAs.BaseEA.BaseEA] = PPGA, pop_size: int = 500, probability_crossover: float = 0.9, probability_mutation: float = 0.3, max_depth: int = 5, max_subtrees: int = 4, prob_terminal: float = 0.5, complexity_scalar: float = 0.5, error_lim: float = 0.001, init_method: str = 'ramped_half_and_half', model_selection_criterion: str = 'min_error', loss_function: str = 'mse', single_obj_generations: int = 10, function_set=('add', 'sub', 'mul', 'div'), terminal_set=None) Bases: :py:obj:`desdeo_problem.surrogatemodels.SurrogateModels.BaseRegressor` Helper class that provides a standard way to create an ABC using inheritance. .. py:method:: fit(X: pandas.DataFrame, y: pandas.DataFrame) .. py:method:: _create_individuals() .. py:method:: _model_performance(trees: LinearNode, X: numpy.ndarray = None, y: numpy.ndarray = None) .. py:method:: predict(X: numpy.ndarray) .. py:method:: select() .. py:method:: add(x, y) :staticmethod: .. py:method:: sub(x, y) :staticmethod: .. py:method:: mul(x, y) :staticmethod: .. py:method:: div(x, y) :staticmethod: .. py:method:: sqrt(x) :staticmethod: .. py:method:: log(x) :staticmethod: .. py:method:: sin(x) :staticmethod: .. py:method:: cos(x) :staticmethod: .. py:method:: tan(x) :staticmethod: .. py:method:: neg(x) :staticmethod: .. py:class:: EvoNN(num_hidden_nodes: int = 20, p_omit: float = 0.2, w_low: float = -5.0, w_high: float = 5.0, activation_function: str = 'sigmoid', loss_function: str = 'mse', training_algorithm: Type[desdeo_emo.EAs.BaseEA.BaseEA] = PPGA, pop_size: int = 500, model_selection_criterion: str = 'akaike_corrected', recombination_type: str = 'evonn_xover_mutation', crossover_type: str = 'standard', mutation_type: str = 'gaussian') Bases: :py:obj:`desdeo_problem.surrogatemodels.SurrogateModels.BaseRegressor` Helper class that provides a standard way to create an ABC using inheritance. .. py:method:: fit(X: numpy.ndarray, y: numpy.ndarray) .. py:method:: _model_performance(first_layer: numpy.ndarray = None, X: numpy.ndarray = None, y_true: numpy.ndarray = None) .. py:method:: predict(X: numpy.ndarray = None, first_layer: numpy.ndarray = None, training: bool = False) .. py:method:: activate(x) .. py:method:: calculate_linear(previous_layer_output) Calculate the final layer using LLSQ or :param non_linear_layer: Output of the activation function :type non_linear_layer: np.ndarray :returns: * **linear_layer** (*np.ndarray*) -- The optimized weight matrix of the upper part of the network * **predicted_values** (*np.ndarray*) -- The prediction of the model * **training_error** (*float*) -- The model's training error .. py:method:: _create_individuals() .. py:method:: select() .. py:class:: EvoDN2(num_subnets: int = 4, num_subsets: int = 4, max_layers: int = 4, max_nodes: int = 4, p_omit: float = 0.2, w_low: float = -5.0, w_high: float = 5.0, subsets: list = None, activation_function: str = 'sigmoid', loss_function: str = 'mse', training_algorithm: desdeo_emo.EAs.BaseEA.BaseEA = PPGA, pop_size: int = 500, model_selection_criterion: str = 'min_error', verbose: int = 0) Bases: :py:obj:`desdeo_problem.surrogatemodels.SurrogateModels.BaseRegressor` Helper class that provides a standard way to create an ABC using inheritance. .. py:method:: fit(X: numpy.ndarray, y: numpy.ndarray) .. py:method:: _model_performance(individuals: numpy.ndarray = None, X: numpy.ndarray = None, y_true: numpy.ndarray = None) .. py:method:: _feed_forward(subnets, X) .. py:method:: _calculate_linear(previous_layer_output) Calculate the final layer using LLSQ or :param non_linear_layer: Output of the activation function :type non_linear_layer: np.ndarray :returns: * **linear_layer** (*np.ndarray*) -- The optimized weight matrix of the upper part of the network * **predicted_values** (*np.ndarray*) -- The prediction of the model .. py:method:: activate(x) .. py:method:: predict(X) .. py:method:: select() .. py:method:: _create_individuals() .. py:class:: surrogateProblem(performance_evaluator) Bases: :py:obj:`desdeo_problem.problem.ProblemBase` The base class for the problems. All other problem classes should be derived from this. .. attribute:: nadir Nadir values for the problem, initiated = None :type: np.ndarray .. attribute:: ideal Ideal values for the problem, initiated = None :type: np.ndarray .. attribute:: nadir_fitness Fitness values for nadir, initiated = None :type: np.ndarray .. attribute:: ideal_fitness Fitness values for ideal, initiated = None :type: np.ndarray .. attribute:: __n_of_objectives Number of objectives, initiated = 0 :type: int .. attribute:: __n_of_variables Number of variables, initiated = 0 :type: int .. attribute:: __decision_vectors Array of decision variable vectors, initiated = None :type: np.ndarray .. attribute:: __objective_vectors Array of objective variable vectors, initiated = None :type: np.ndarray .. py:method:: evaluate(model_parameters, use_surrogates=False) Abstract method to evaluate problem. Evaluates the problem using an ensemble of input vectors. Uses surrogate models if available. Otherwise, it uses the true evaluator. :param decision_vectors: An array of decision variable :type decision_vectors: np.ndarray :param input vectors.: :param use_surrogate: A bool to control whether to use the true, potentially :type use_surrogate: bool :param expensive function or a surrogate model to evaluate the objectives.: :returns: Dict with the following keys: 'objectives' (np.ndarray): The objective function values for each input vector. 'constraints' (Union[np.ndarray, None]): The constraint values of the problem corresponding each input vector. 'fitness' (np.ndarray): Equal to objective values if objective is to be minimized. Multiplied by (-1) if objective to be maximized. 'uncertainity' (Union[np.ndarray, None]): The uncertainity in the objective values. :rtype: (Dict) .. py:method:: evaluate_constraint_values() Abstract method to evaluate constraint values. Evaluate just the constraint function values using the attributes decision_vectors and objective_vectors .. note:: Currently not supported by ScalarMOProblem .. py:method:: get_variable_bounds() Abstract method to get variable bounds .. py:method:: get_objective_names()