#!/usr/bin/env python
# Created by "Thieu" at 14:52, 26/05/2023 ----------%
# Email: nguyenthieu2102@gmail.com %
# Github: https://github.com/thieu1995 %
# --------------------------------------------------%
import numpy as np
from mealpy.optimizer import Optimizer
from mealpy import *
[docs]class FeatureSelectionProblem(Problem):
def __init__(self, lb, ub, minmax, data=None, estimator=None,
transfer_func=None, obj_name=None,
metric_class=None, fit_weights=(0.9, 0.1), fit_sign=1, obj_paras=None,
name="Feature Selection Problem", **kwargs):
## data is assigned first because when initialize the Problem class, we need to check the output of fitness
self.data = data
self.estimator = estimator
self.transfer_func = transfer_func
self.obj_name = obj_name
self.metric_class = metric_class
self.fit_weights = fit_weights
self.fit_sign = fit_sign
self.obj_paras = obj_paras
self.name = name
super().__init__(lb, ub, minmax, **kwargs)
[docs] def amend_position(self, position=None, lb=None, ub=None):
position = self.transfer_func(position)
cons = np.random.uniform(0, 1, len(position))
x = np.where(cons < position, 1, 0)
if np.sum(x) == 0:
x[np.random.randint(0, len(x))] = 1
return x
[docs] def fit_func(self, solution):
cols = np.flatnonzero(solution)
self.estimator.fit(self.data.X_train[:, cols], self.data.y_train)
y_valid_pred = self.estimator.predict(self.data.X_test[:, cols])
evaluator = self.metric_class(self.data.y_test, y_valid_pred)
obj = evaluator.get_metric_by_name(self.obj_name, paras=self.obj_paras)[self.obj_name]
fitness = self.fit_weights[0]*obj + self.fit_sign * self.fit_weights[1]*(np.sum(solution)/self.n_dims)
return [fitness, obj]