#!/usr/bin/env python
# -*- coding: utf-8 -*-
# @Author : qichun tang
# @Date : 2020-12-14
# @Contact : qichun.tang@bupt.edu.cn
from skopt.learning.forest import ExtraTreesRegressor, RandomForestRegressor
from skopt.learning.gbrt import GradientBoostingQuantileRegressor
from ultraopt.optimizer.bo.sampling_sort_opt import SamplingSortOptimizer
from ultraopt.utils.config_transformer import ConfigTransformer
[文档]class ForestOptimizer(SamplingSortOptimizer):
def __init__(
self,
# model related
forest_type="ET",
n_estimators=10, max_depth=None, max_features="auto",
# several hyper-parameters
use_local_search=False, loss_transformer="log_scaled",
min_points_in_model=20, n_samples=5000,
acq_func="EI", xi=0,
# model related (trival)
min_samples_leaf=1, min_weight_fraction_leaf=0, min_samples_split=2,
max_leaf_nodes=None, min_impurity_decrease=0, bootstrap=True, oob_score=False,
n_jobs=1, min_variance=0,
):
if forest_type == "ET":
forest_klass = ExtraTreesRegressor
elif forest_type == "RF":
forest_klass = RandomForestRegressor
else:
raise ValueError(
f"forest_type should be 'ET' or 'RF', means 'ExtraTrees' and 'RandomForest', respectively. ")
super(ForestOptimizer, self).__init__(
epm=forest_klass(
n_estimators=n_estimators,
max_depth=max_depth,
min_samples_split=min_samples_split,
min_samples_leaf=min_samples_leaf,
min_weight_fraction_leaf=min_weight_fraction_leaf,
max_features=max_features,
max_leaf_nodes=max_leaf_nodes,
min_impurity_decrease=min_impurity_decrease,
bootstrap=bootstrap,
oob_score=oob_score,
n_jobs=n_jobs,
min_variance=min_variance,
# random_state=random_state,
),
config_transformer=ConfigTransformer(impute=-1, encoder=None),
use_local_search=use_local_search,
loss_transformer=loss_transformer,
min_points_in_model=min_points_in_model,
n_samples=n_samples,
acq_func=acq_func,
xi=xi
)
[文档] def initialize(self, config_space, budgets=(1,), random_state=42, initial_points=None, budget2obvs=None):
super(ForestOptimizer, self).initialize(config_space, budgets, random_state, initial_points,
budget2obvs)
self.epm.set_params(random_state=random_state)
[文档]class GBRTOptimizer(SamplingSortOptimizer):
def __init__(
self,
# model related
n_jobs=1,
# several hyper-parameters
use_local_search=False, loss_transformer="log_scaled",
min_points_in_model=20, n_samples=5000,
acq_func="EI", xi=0
):
super(GBRTOptimizer, self).__init__(
epm=GradientBoostingQuantileRegressor(
n_jobs=n_jobs,
),
config_transformer=ConfigTransformer(impute=-1, encoder=None),
use_local_search=use_local_search,
loss_transformer=loss_transformer,
min_points_in_model=min_points_in_model,
n_samples=n_samples,
acq_func=acq_func,
xi=xi
)
[文档] def initialize(self, config_space, budgets=(1,), random_state=42, initial_points=None, budget2obvs=None):
super(GBRTOptimizer, self).initialize(config_space, budgets, random_state, initial_points,
budget2obvs)
self.epm.set_params(random_state=random_state)
