Python实现LSTM多变量多步滚动预测代码

以下是Python实现LSTM多变量多步滚动预测的代码示例：

import numpy as np
import pandas as pd
from sklearn.preprocessing import MinMaxScaler
from keras.models import Sequential
from keras.layers import Dense, LSTM

# 加载数据
df = pd.read_csv('data.csv', header=0, index_col=0)

# 数据预处理
scaler = MinMaxScaler(feature_range=(0, 1))
scaled_data = scaler.fit_transform(df.values)

# 定义函数将数据转换为多时步数据集
def create_dataset(dataset, time_steps=1):
    X, y = [], []
    for i in range(len(dataset) - time_steps):
        X.append(dataset[i:(i + time_steps), :])
        y.append(dataset[i + time_steps, 0])
    return np.array(X), np.array(y)

# 设置时步和训练/测试集大小
time_steps = 3
train_size = int(len(scaled_data) * 0.75)
test_size = len(scaled_data) - train_size

# 创建训练/测试数据集
train_data = scaled_data[0:train_size, :]
test_data = scaled_data[train_size:len(scaled_data), :]

# 转换数据为多时步数据集
X_train, y_train = create_dataset(train_data, time_steps)
X_test, y_test = create_dataset(test_data, time_steps)

# 定义模型
model = Sequential()
model.add(LSTM(units=50, return_sequences=True, input_shape=(X_train.shape[1], X_train.shape[2])))
model.add(LSTM(units=50))
model.add(Dense(1))
model.compile(loss='mean_squared_error', optimizer='adam')

# 训练模型
model.fit(X_train, y_train, epochs=100, batch_size=32, validation_data=(X_test, y_test), verbose=2)

# 进行预测
train_predict = model.predict(X_train)
test_predict = model.predict(X_test)

# 反向转换预测结果
train_predict = scaler.inverse_transform(train_predict)
y_train = scaler.inverse_transform([y_train])
test_predict = scaler.inverse_transform(test_predict)
y_test = scaler.inverse_transform([y_test])

# 计算均方根误差
train_score = np.sqrt(mean_squared_error(y_train[0], train_predict[:, 0]))
print('Train RMSE: %.2f' % (train_score))
test_score = np.sqrt(mean_squared_error(y_test[0], test_predict[:, 0]))
print('Test RMSE: %.2f' % (test_score))

需要注意的是，此代码需要导入以下库：

- numpy
- pandas
- sklearn
- keras

其中，数据集应为csv格式，模型使用LSTM，使用均方根误差作为模型评估指标。