波士顿房价数据集中，遗传算法和粒子群算法优化随机森林的代码

以下是使用遗传算法和粒子群算法优化随机森林的代码：

使用遗传算法优化随机森林：

from sklearn.ensemble import RandomForestRegressor
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
from geneticalgorithm import geneticalgorithm as ga

# 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)

# 定义优化函数
def fitness_function(X):
    n_estimators = int(X[0])
    max_depth = int(X[1])
    max_features = int(X[2])
    min_samples_split = int(X[3])
    min_samples_leaf = int(X[4])
    
    # 训练模型
    rf = RandomForestRegressor(n_estimators=n_estimators, 
                                max_depth=max_depth,
                                max_features=max_features,
                                min_samples_split=min_samples_split,
                                min_samples_leaf=min_samples_leaf,
                                random_state=42)
    rf.fit(X_train, y_train)
    
    # 计算均方误差
    y_pred = rf.predict(X_test)
    mse = mean_squared_error(y_test, y_pred)
    
    return -mse

# 定义参数范围
varbound = np.array([[10, 200], # n_estimators
                     [1, 50],   # max_depth
                     [1, 13],   # max_features
                     [2, 50],   # min_samples_split
                     [1, 50]]) # min_samples_leaf

# 使用遗传算法进行优化
model = ga(function=fitness_function, dimension=5, variable_type='int', variable_boundaries=varbound)
model.run()

使用粒子群算法优化随机森林：

from sklearn.ensemble import RandomForestRegressor
from sklearn.datasets import load_boston
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error
from pyswarm import pso

# 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)

# 定义优化函数
def fitness_function(X):
    n_estimators = int(X[0])
    max_depth = int(X[1])
    max_features = int(X[2])
    min_samples_split = int(X[3])
    min_samples_leaf = int(X[4])
    
    # 训练模型
    rf = RandomForestRegressor(n_estimators=n_estimators, 
                                max_depth=max_depth,
                                max_features=max_features,
                                min_samples_split=min_samples_split,
                                min_samples_leaf=min_samples_leaf,
                                random_state=42)
    rf.fit(X_train, y_train)
    
    # 计算均方误差
    y_pred = rf.predict(X_test)
    mse = mean_squared_error(y_test, y_pred)
    
    return mse

# 定义参数范围
lb = [10, 1, 1, 2, 1]
ub = [200, 50, 13, 50, 50]

# 使用粒子群算法进行优化
xopt, fopt = pso(fitness_function, lb, ub, swarmsize=100, maxiter=100)