用tensorflow实现lstm预测

1. 导入所需库

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

2. 加载数据

df = pd.read_csv('data.csv')

3. 数据预处理

# 将数据转换为numpy数组
data = df.values

# 将数据缩放到0到1之间
scaler = MinMaxScaler(feature_range=(0, 1))
scaled_data = scaler.fit_transform(data)

# 将数据拆分为训练集和测试集
train_size = int(len(scaled_data) * 0.8)
test_size = len(scaled_data) - train_size
train_data = scaled_data[0:train_size, :]
test_data = scaled_data[train_size:len(scaled_data), :]

# 将数据转换为x_train和y_train
def create_dataset(dataset, time_step=1):
    data_x, data_y = [], []
    for i in range(len(dataset) - time_step - 1):
        a = dataset[i:(i + time_step), 0]
        data_x.append(a)
        data_y.append(dataset[i + time_step, 0])
    return np.array(data_x), np.array(data_y)

time_step = 100
x_train, y_train = create_dataset(train_data, time_step)
x_test, y_test = create_dataset(test_data, time_step)

# 将数据转换为3D格式
x_train = x_train.reshape(x_train.shape[0], x_train.shape[1], 1)
x_test = x_test.reshape(x_test.shape[0], x_test.shape[1], 1)

4. 构建模型

model = Sequential()
model.add(LSTM(50, return_sequences=True, input_shape=(x_train.shape[1], 1)))
model.add(LSTM(50, return_sequences=True))
model.add(LSTM(50))
model.add(Dense(1))

model.compile(optimizer='adam', loss='mean_squared_error')

5. 模型训练

model.fit(x_train, y_train, validation_data=(x_test, y_test), epochs=100, batch_size=64, verbose=1)

6. 模型预测

train_predict = model.predict(x_train)
test_predict = model.predict(x_test)

# 将预测值反缩放
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rmse
train_score = np.sqrt(np.mean(np.power(y_train - train_predict, 2)))
test_score = np.sqrt(np.mean(np.power(y_test - test_predict, 2)))

7. 可视化结果

import matplotlib.pyplot as plt

# 绘制训练集预测结果和实际值
plt.plot(y_train[0])
plt.plot(train_predict)
plt.title("Train Predictions")
plt.xlabel("Time")
plt.ylabel("Value")
plt.legend(["Actual", "Predicted"])
plt.show()

# 绘制测试集预测结果和实际值
plt.plot(y_test[0])
plt.plot(test_predict)
plt.title("Test Predictions")
plt.xlabel("Time")
plt.ylabel("Value")
plt.legend(["Actual", "Predicted"])
plt.show()