使用 Python 的 LSTM 进行股市预测

import numpy as np
from keras.models import Sequential
from keras.layers import LSTM, Dense
from sklearn.preprocessing import MinMaxScaler
from sklearn.metrics import mean_squared_error
 
# 示例股票价格数据
def load_data(filename):
    with open(filename, 'r') as f:
        data = f.read()
    all_data = data.split('\n')
    stock_data = np.array(all_data).astype(np.float)
    return stock_data
 
# 将数据转换为LSTM模型需要的形状
def prepare_data(stock_data, seq_length, normalize_window=True):
    X, y = [], []
    for i in range(len(stock_data) - seq_length):
        X.append(stock_data[i:i + seq_length])
        y.append(stock_data[i + seq_length])
    X = np.array(X).reshape((X.shape[0], seq_length, 1))
    y = np.array(y).reshape((y.shape[0], 1))
    if normalize_window:
        scaler = MinMaxScaler(feature_range=(0, 1))
        X_scaled = scaler.fit_transform(X.reshape((X.shape[0], X.shape[2])))
        X = X_scaled.reshape((X.shape[0], X.shape[1], X.shape[2]))
    return X, y
 
# 模型训练和评估函数
def train_and_evaluate_model(X_train, y_train, X_test, y_test, seq_length, n_features, batch_size, epochs):
    model = Sequential()
    model.add(LSTM(50, input_shape=(seq_length, n_features)))
    model.add(Dense(1))
    model.compile(optimizer='adam', loss='mse')
    model.fit(X_train, y_train, epochs=epochs, batch_size=batch_size, verbose=2)
    # 评估模型
    _, ax = plt.subplots(figsize=(12, 4))
    test_predict = model.predict(X_test)
    ax.plot(test_predict, color='blue', label='Prediction')
    ax.plot(y_test, color='red', label='Real Value')
    ax.legend(loc='upper left')
    ax.set_xlabel('Time')
    ax.set_ylabel('Stock Price')
    ax.set_title('Model Prediction vs. Real Stock Price')
    plt.show()
    # 计算平均平方误差
    mse = mean_squared_error(y_test, test_predict)
    return mse
 
# 示例使用
seq_length = 30  # 序列长度
batch_size = 1  # 批量大小
epochs = 500  # 训练轮数
n_features = 1  # 特征数量（单变量数据）
 
# 加载数据
stock_data = load_data('stock_data.txt')
# 准备数据
X_train, y_train = prepare_data(stock_data[:580], seq_length, normalize_window=True)
X_test, y_test = prepare_data(stock_data[580:], seq_length, normalize_window=True)
# 训练和评估模型
mse = train_and_evaluate_model(X_train, y_train, X_test, y_test, seq_length, n_features, batch_size, epochs)
print(f"Mean Squared Error: {mse}")

这段代码展示了如何加载股票价格数据，将其转换