Source code for mealpy.human_based.LCO

#!/usr/bin/env python
# Created by "Thieu" at 11:16, 18/03/2020 ----------%
#       Email: nguyenthieu2102@gmail.com            %
#       Github: https://github.com/thieu1995        %
# --------------------------------------------------%

import numpy as np
from mealpy.optimizer import Optimizer


[docs]class OriginalLCO(Optimizer): """ The original version of: Life Choice-based Optimization (LCO) Links: 1. https://doi.org/10.1007/s00500-019-04443-z Hyper-parameters should fine-tune in approximate range to get faster convergence toward the global optimum: + r1 (float): [1.5, 4], coefficient factor, default = 2.35 Examples ~~~~~~~~ >>> import numpy as np >>> from mealpy import FloatVar, LCO >>> >>> def objective_function(solution): >>> return np.sum(solution**2) >>> >>> problem_dict = { >>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"), >>> "minmax": "min", >>> "obj_func": objective_function >>> } >>> >>> model = LCO.OriginalLCO(epoch=1000, pop_size=50, r1 = 2.35) >>> g_best = model.solve(problem_dict) >>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}") >>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}") References ~~~~~~~~~~ [1] Khatri, A., Gaba, A., Rana, K.P.S. and Kumar, V., 2020. A novel life choice-based optimizer. Soft Computing, 24(12), pp.9121-9141. """ def __init__(self, epoch: int = 10000, pop_size: int = 100, r1: float = 2.35, **kwargs: object) -> None: """ Args: epoch (int): maximum number of iterations, default = 10000 pop_size (int): number of population size, default = 100 r1 (float): coefficient factor """ super().__init__(**kwargs) self.epoch = self.validator.check_int("epoch", epoch, [1, 100000]) self.pop_size = self.validator.check_int("pop_size", pop_size, [5, 10000]) self.r1 = self.validator.check_float("r1", r1, [1.0, 3.0]) self.set_parameters(["epoch", "pop_size", "r1"]) self.n_agents = int(np.ceil(np.sqrt(self.pop_size))) self.sort_flag = True
[docs] def evolve(self, epoch): """ The main operations (equations) of algorithm. Inherit from Optimizer class Args: epoch (int): The current iteration """ pop_new = [] for idx in range(0, self.pop_size): prob = self.generator.random() if prob > 0.875: # Update using Eq. 1, update from n best position temp = np.array([self.generator.random() * self.pop[j].solution for j in range(0, self.n_agents)]) temp = np.mean(temp, axis=0) elif prob < 0.7: # Update using Eq. 2-6 f1 = 1 - epoch / self.epoch f2 = 1 - f1 prev_pos = self.g_best.solution if idx == 0 else self.pop[idx-1].solution best_diff = f1 * self.r1 * (self.g_best.solution - self.pop[idx].solution) better_diff = f2 * self.r1 * (prev_pos - self.pop[idx].solution) temp = self.pop[idx].solution + self.generator.random() * better_diff + self.generator.random() * best_diff else: temp = self.problem.ub - (self.pop[idx].solution - self.problem.lb) * self.generator.random() pos_new = self.correct_solution(temp) agent = self.generate_empty_agent(pos_new) pop_new.append(agent) if self.mode not in self.AVAILABLE_MODES: agent.target = self.get_target(pos_new) self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax) if self.mode in self.AVAILABLE_MODES: pop_new = self.update_target_for_population(pop_new) self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)
[docs]class DevLCO(OriginalLCO): """ The developed version: Life Choice-based Optimization (LCO) Notes: + The flow is changed with if else statement. Hyper-parameters should fine-tune in approximate range to get faster convergence toward the global optimum: + r1 (float): [1.5, 4], coefficient factor, default = 2.35 Examples ~~~~~~~~ >>> import numpy as np >>> from mealpy import FloatVar, LCO >>> >>> def objective_function(solution): >>> return np.sum(solution**2) >>> >>> problem_dict = { >>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"), >>> "minmax": "min", >>> "obj_func": objective_function >>> } >>> >>> model = LCO.DevLCO(epoch=1000, pop_size=50, r1 = 2.35) >>> g_best = model.solve(problem_dict) >>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}") >>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}") """ def __init__(self, epoch: int = 10000, pop_size: int = 100, r1: float = 2.35, **kwargs: object) -> None: """ Args: epoch (int): maximum number of iterations, default = 10000 pop_size (int): number of population size, default = 100 r1 (float): coefficient factor """ super().__init__(epoch, pop_size, r1, **kwargs)
[docs] def evolve(self, epoch): """ The main operations (equations) of algorithm. Inherit from Optimizer class Args: epoch (int): The current iteration """ # epoch: current chance, self.epoch: number of chances pop_new = [] for idx in range(0, self.pop_size): prob = self.generator.random() if prob > 0.875: # Update using Eq. 1, update from n best position temp = np.array([self.generator.random() * self.pop[j].solution for j in range(0, self.n_agents)]) temp = np.mean(temp, axis=0) elif prob < 0.7: # Update using Eq. 2-6 f = epoch / self.epoch if idx != 0: better_diff = f * self.r1 * (self.pop[idx - 1].solution - self.pop[idx].solution) else: better_diff = f * self.r1 * (self.g_best.solution - self.pop[idx].solution) best_diff = (1 - f) * self.r1 * (self.pop[0].solution - self.pop[idx].solution) temp = self.pop[idx].solution + self.generator.random() * better_diff + self.generator.random() * best_diff else: temp = self.problem.generate_solution() pos_new = self.correct_solution(temp) agent = self.generate_empty_agent(pos_new) pop_new.append(agent) if self.mode not in self.AVAILABLE_MODES: agent.target = self.get_target(pos_new) self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax) if self.mode in self.AVAILABLE_MODES: pop_new = self.update_target_for_population(pop_new) self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax)
[docs]class ImprovedLCO(Optimizer): """ The improved version: Life Choice-based Optimization (ILCO) Notes: + The flow of the original LCO is kept. + Gaussian distribution and mutation mechanism are added + R1 parameter is removed Examples ~~~~~~~~ >>> import numpy as np >>> from mealpy import FloatVar, LCO >>> >>> def objective_function(solution): >>> return np.sum(solution**2) >>> >>> problem_dict = { >>> "bounds": FloatVar(n_vars=30, lb=(-10.,) * 30, ub=(10.,) * 30, name="delta"), >>> "minmax": "min", >>> "obj_func": objective_function >>> } >>> >>> model = LCO.ImprovedLCO(epoch=1000, pop_size=50) >>> g_best = model.solve(problem_dict) >>> print(f"Solution: {g_best.solution}, Fitness: {g_best.target.fitness}") >>> print(f"Solution: {model.g_best.solution}, Fitness: {model.g_best.target.fitness}") """ def __init__(self, epoch: int = 10000, pop_size: int = 100, **kwargs: object) -> None: """ Args: epoch (int): maximum number of iterations, default = 10000 pop_size (int): number of population size, default = 100 """ super().__init__(**kwargs) self.epoch = self.validator.check_int("epoch", epoch, [1, 100000]) self.pop_size = self.validator.check_int("pop_size", pop_size, [5, 10000]) self.set_parameters(["epoch", "pop_size"]) self.pop_len = int(self.pop_size / 2) self.sort_flag = True
[docs] def evolve(self, epoch): """ The main operations (equations) of algorithm. Inherit from Optimizer class Args: epoch (int): The current iteration """ # epoch: current chance, self.epoch: number of chances pop_new = [] for idx in range(0, self.pop_size): rand = self.generator.random() if rand > 0.875: # Update using Eq. 1, update from n best position n = int(np.ceil(np.sqrt(self.pop_size))) pos_new = np.array([self.generator.random() * self.pop[j].solution for j in range(0, n)]) pos_new = np.mean(pos_new, axis=0) elif rand < 0.7: # Update using Eq. 2-6 f = epoch / self.epoch if idx != 0: better_diff = f * self.generator.random() * (self.pop[idx - 1].solution - self.pop[idx].solution) else: better_diff = f * self.generator.random() * (self.g_best.solution - self.pop[idx].solution) best_diff = (1 - f) * self.generator.random() * (self.pop[0].solution - self.pop[idx].solution) pos_new = self.pop[idx].solution + better_diff + best_diff else: pos_new = self.problem.ub - (self.pop[idx].solution - self.problem.lb) * self.generator.random() pos_new = self.correct_solution(pos_new) agent = self.generate_empty_agent(pos_new) pop_new.append(agent) if self.mode not in self.AVAILABLE_MODES: agent.target = self.get_target(pos_new) self.pop[idx] = self.get_better_agent(agent, self.pop[idx], self.problem.minmax) if self.mode in self.AVAILABLE_MODES: pop_new = self.update_target_for_population(pop_new) self.pop = self.greedy_selection_population(self.pop, pop_new, self.problem.minmax) ## Sort the updated population based on fitness pop = self.get_sorted_population(self.pop, self.problem.minmax) local_best = pop[0].copy() pop_s1 = [agent.copy() for agent in pop[:self.pop_len]] pop_s2 = [agent.copy() for agent in pop[self.pop_len:]] ## Mutation scheme pop_child1 = [] for idx in range(0, self.pop_len): pos_new = pop_s1[idx].solution + self.generator.normal(0, 1, self.problem.n_dims) * pop_s1[idx].solution pos_new = self.correct_solution(pos_new) agent = self.generate_empty_agent(pos_new) pop_child1.append(agent) if self.mode not in self.AVAILABLE_MODES: agent.target = self.get_target(pos_new) pop_s1[idx] = self.get_better_agent(agent, pop_s1[idx], self.problem.minmax) if self.mode in self.AVAILABLE_MODES: pop_child1 = self.update_target_for_population(pop_child1) pop_s1 = self.greedy_selection_population(pop_s1, pop_child1, self.problem.minmax) ## Search Mechanism pos_s1_list = [agent.solution for agent in pop_s1] pos_s1_mean = np.mean(pos_s1_list, axis=0) pop_child2 = [] for idx in range(0, self.pop_len): pos_new = local_best.solution + self.generator.uniform(0, 1) * pos_s1_mean * (epoch / self.epoch) pos_new = self.correct_solution(pos_new) agent = self.generate_empty_agent(pos_new) pop_child2.append(agent) if self.mode not in self.AVAILABLE_MODES: agent.target = self.get_target(pos_new) pop_s2[idx] = self.get_better_agent(pop_s2[idx], agent, self.problem.minmax) if self.mode in self.AVAILABLE_MODES: pop_child2 = self.update_target_for_population(pop_s2) pop_s2 = self.greedy_selection_population(pop_s2, pop_child2, self.problem.minmax) ## Construct a new population self.pop = pop_s1 + pop_s2