波士顿房价中遗传算法优化lightgbm的代码

以下是一个使用遗传算法优化LightGBM模型的波士顿房价预测代码示例：

import pandas as pd
import numpy as np
from sklearn.datasets import load_boston
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split
from lightgbm import LGBMRegressor
from deap import creator, base, tools, algorithms

# 加载数据集
boston = load_boston()
X, y = boston.data, boston.target

# 划分训练集和测试集
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

# 定义适应度函数
def eval_func(params):
    # 设置LightGBM的参数
    lgb_reg = LGBMRegressor(
        n_estimators=params[0],
        learning_rate=params[1],
        num_leaves=params[2],
        max_depth=params[3],
        min_child_samples=params[4],
        subsample=params[5],
        colsample_bytree=params[6]
    )
    # 训练模型并预测
    lgb_reg.fit(X_train, y_train)
    y_pred = lgb_reg.predict(X_test)
    # 计算均方误差
    mse = mean_squared_error(y_test, y_pred)
    return mse,

# 遗传算法设置
creator.create("FitnessMin", base.Fitness, weights=(-1.0,))
creator.create("Individual", list, fitness=creator.FitnessMin)

toolbox = base.Toolbox()
toolbox.register("attr_int", np.random.randint, 50, 500)
toolbox.register("attr_float", np.random.uniform, 0.001, 0.1)
toolbox.register("individual", tools.initCycle, creator.Individual,
                 (toolbox.attr_int, toolbox.attr_float), n=7)
toolbox.register("population", tools.initRepeat, list, toolbox.individual)
toolbox.register("mate", tools.cxTwoPoint)
toolbox.register("mutate", tools.mutGaussian, mu=0, sigma=0.1, indpb=0.1)
toolbox.register("select", tools.selTournament, tournsize=3)
toolbox.register("evaluate", eval_func)

# 运行遗传算法
def main():
    pop_size = 50
    num_generations = 10
    crossover_prob = 0.5
    mutation_prob = 0.2

    pop = toolbox.population(n=pop_size)
    hof = tools.HallOfFame(1)

    stats = tools.Statistics(lambda ind: ind.fitness.values)
    stats.register("min", np.min)
    stats.register("avg", np.mean)

    pop, log = algorithms.eaSimple(pop, toolbox, cxpb=crossover_prob, mutpb=mutation_prob,
                                   ngen=num_generations, stats=stats, halloffame=hof, verbose=True)

    best_ind = hof[0]
    print("Best individual is ", best_ind)
    print("Best fitness is ", best_ind.fitness.values[0])
    
    # 使用最佳参数训练模型并预测
    lgb_reg = LGBMRegressor(
        n_estimators=best_ind[0],
        learning_rate=best_ind[1],
        num_leaves=best_ind[2],
        max_depth=best_ind[3],
        min_child_samples=best_ind[4],
        subsample=best_ind[5],
        colsample_bytree=best_ind[6]
    )
    lgb_reg.fit(X_train, y_train)
    y_pred = lgb_reg.predict(X_test)
    mse = mean_squared_error(y_test, y_pred)
    print("MSE on test set with best individual: ", mse)

if __name__ == "__main__":
    main()

代码中使用了DEAP库来实现遗传算法，通过定义适应度函数来评估每个个体的表现。使用`LGBMRegressor`构建LightGBM模型，并使用均方误差作为适应度函数的评价指标。在遗传算法中，使用了7个超参数进行优化，包括n_estimators、learning_rate、num_leaves、max_depth、min_child_samples、subsample和colsample_bytree。最后，使用最佳超参数训练模型并预测波士顿房价，并计算预测误差。