import numpy as np
from mlrose import mlrose
edges = [(0, 1), (1, 2), (0, 2), (1, 3), (2, 3), (3, 4)]
fitness = mlrose.MaxKColor(edges)
problem = mlrose.DiscreteOpt(length=5,
fitness_fn=fitness,
maximize=False,
max_val=2)
rhc = mlrose.RHCRunner(
problem=problem,
experiment_name="RHC_final",
output_directory=
"/Users/matthieudivet/Desktop/GaTech/Classes/ML/Assignments/Randomized_optimization/k_color_problem",
seed=None,
iteration_list=2**np.arange(14),
max_attempts=1000,
restart_list=[0])
rhc_run_stats, rhc_run_curves = rhc.run()
sa = mlrose.SARunner(
problem=problem,
experiment_name="SA_final",
output_directory=
"/Users/matthieudivet/Desktop/GaTech/Classes/ML/Assignments/Randomized_optimization/k_color_problem",
seed=None,
iteration_list=2**np.arange(14),
max_attempts=1000,
temperature_list=[250],
decay_list=[mlrose.ExpDecay])
from mlrose import mlrose
import numpy as np
import matplotlib.pyplot as plt
import time
from util import plot_fitness_curve
np.random.seed(seed=0)
# Initialize fitness function object
fitness_fn = mlrose.ContinuousPeaks(t_pct=0.1)
state = np.random.randint(low=0, high=2, size=1000)
# Define optimization problem object
problem_fit = mlrose.DiscreteOpt(length=len(state),
fitness_fn=fitness_fn,
maximize=True,
max_val=2)
iterations = [1, 10, 100, 1000, 10000]
# ________________________________________________________________________________________________________________________
times_ga = []
fitns_ga = []
# Solve problem using the genetic algorithm
for i in iterations:
start = time.time()
best_state, best_fitness, c = mlrose.genetic_alg(problem_fit,
pop_size=200,
mutation_prob=0.1,
curve=True,
max_iters=i,
random_state=3)
import pickle
ps = 32
mimic_maxs = []
mimic_its = []
for i in range(5):
final_results_maxs = []
final_results_it = []
for p in range(2, ps + 1):
print(p)
fitness = mlrose.FlipFlop()
istate = np.array(np.zeros(p), dtype=int)
problem = mlrose.DiscreteOpt(length=p,
fitness_fn=fitness,
maximize=True,
max_val=2)
schedule = mlrose.ExpDecay(init_temp=0.5,
exp_const=0.005,
min_temp=0.001)
best_state, best_fitness, fitness_curve = mlrose.mimic(
problem,
max_attempts=500,
pop_size=500,
keep_pct=0.3,
max_iters=1000,
curve=True)
# Define alternative N-Queens fitness function for maximization problem
import numpy as np
import mlrose.mlrose as mlrose
import pandas as pd
import pickle
ps = 16
results = []
for i in range(10,1000,25):
n_queen = ps
fitness_1 = mlrose.Queens()
problem = mlrose.DiscreteOpt(length = n_queen, fitness_fn = fitness_1,
maximize = False, max_val = n_queen)
istate = np.array(np.zeros(ps), dtype=int)
schedule = mlrose.ExpDecay(init_temp=0.5, exp_const=0.005, min_temp=0.001)
best_state, best_fitness,fitness_curve = mlrose.mimic(problem,max_attempts =500,pop_size=i,keep_pct=0.3, max_iters = 500,random_state = 1,curve=True)
# Define alternative N-Queens fitness function for maximization problem
# Initialize custom fitness function object
# Define initial state
print('The best state found is: ', best_state)
print('The fitness at the best state is: ', best_fitness)