如何输入4个值（‘开盘价’，‘最高价’，‘最低价’，‘成交总量’）来预测未来x天的相同4个值？

Question

我正在做一个机器学习模型，这个模型需要输入以下股票数据来进行预测：
'开盘价'、'最高价'、'最低价'、'总交易量'

我觉得我的代码有问题。怎样才能正确地预测未来x天的股票价格呢？

运行代码的前提条件：

!git clone https://github.com/NSEDownload/NSEDownload
# installation of NSEDownload library
!pip3 install NSEDownload/dist/*

以下是我的代码：

from sklearn.model_selection import TimeSeriesSplit
from sklearn.preprocessing import MinMaxScaler
import pandas as pd
import numpy as np
from keras.models import Sequential
from keras.layers import LSTM, Dense
import matplotlib.pyplot as plt
import matplotlib.dates as mdates
from NSEDownload import stocks

stock_name = 'TCS'

# Get stock data
df = stocks.get_data(stock_symbol=stock_name, full_data=True)
df.index = pd.to_datetime(df.index)
df.to_csv(f'{stock_name}.csv')

# Number of past days to use
n_days = 10

print("Dataframe shape (rows, columns): ", df.shape)
print("Are any null values present? ", df.isnull().values.any())

# set the target variable
training_df = pd.DataFrame(df['Last Price'])
#Selecting the Features
features = ['Open Price', 'High Price', 'Low Price', 'Total Traded Quantity']

# scaling (normalization)
scaler = MinMaxScaler()
training_df_transform = scaler.fit_transform(df[features])
training_df_transform= pd.DataFrame(columns=features, data=training_df_transform, index=df.index)

timesplit= TimeSeriesSplit(n_splits=10)
for train_index, test_index in timesplit.split(df.index):
    X_train, X_test = training_df_transform.iloc[:len(train_index)], training_df_transform.iloc[len(train_index): (len(train_index) + len(test_index))]
    y_train, y_test = training_df.iloc[:len(train_index)].values.ravel(), training_df.iloc[len(train_index): (len(train_index) + len(test_index))].values.ravel()

# Create sequences of length n_days
X_train = np.array([X_train[i : i + n_days] for i in range(len(X_train) - n_days)])
X_test = np.array([X_test[i : i + n_days] for i in range(len(X_test) - n_days)])

# Adjust y_train and y_test to match the new X_train and X_test
y_train = y_train[n_days:]
y_test = y_test[n_days:]

# Reshape X_train and X_test to match the shape the model expects
X_train = X_train.reshape((X_train.shape[0], X_train.shape[1], X_train.shape[2]))
X_test = X_test.reshape((X_test.shape[0], X_test.shape[1], X_test.shape[2]))

# Define and train the model
model = Sequential()
model.add(LSTM(32, input_shape=(n_days, X_train.shape[2]), activation='relu', return_sequences=False))
model.add(Dense(4))  # Change this to match the number of output features
model.compile(loss='mean_squared_error', optimizer='adam')
model.fit(X_train, y_train, epochs=50, batch_size=8, verbose=1, shuffle=False)

# Predict
y_pred = model.predict(X_test)
# Plotting code
plt.plot(df.index[-len(y_test):], y_test, label='Actual Close Price')
plt.plot(df.index[-len(y_test):], y_pred[:, 0], label='Predicted Close Price')
# Set major locator and formatter for x-axis
plt.gca().xaxis.set_major_locator(mdates.MonthLocator(interval=2))
plt.gca().xaxis.set_major_formatter(mdates.DateFormatter('%b %Y'))
plt.title('Testing the Model Prediction for ' + stock_name)
plt.xlabel('Month')
plt.ylabel('Stock Price (Rs.)')
plt.legend()
# Rotate x-axis labels for better readability
plt.xticks(rotation=45)
plt.tight_layout()
plt.show()