Blog

mport numpy as np
import pandas as pd
from sklearn.ensemble import RandomForestRegressor
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import mean_absolute_error
from sklearn.model_selection import GridSearchCV
import matplotlib.pyplot as plt
import seaborn as sns
import talib

Load historical market data
data = pd.read_csv("market_data.csv")
data = data[['date', 'open', 'high', 'low', 'close']]

Create feature set for machine learning model
data['HL_PCT'] = (data['high'] - data['low']) / data['low'] * 100.0
data['PCT_change'] = (data['close'] - data['open']) / data['open'] * 100.0
data = data[['date', 'open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']]

Split data into train and test sets
train_data = data[:int(0.8len(data))]
test_data = data[int(0.8len(data)):]

Scale feature data for better model performance
scaler = StandardScaler()
scaler.fit(train_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']])
train_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']] = scaler.transform(train_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']])
test_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']] = scaler.transform(test_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']])

Create machine learning model
rf = RandomForestRegressor()
param_grid = {'n_estimators': [10, 50, 100], 'max_depth': [5, 10, 20]}
grid = GridSearchCV(estimator=rf, param_grid=param_grid, cv=3, n_jobs=-1)
grid.fit(train_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']], train_data['close'])

Make predictions on test data
predictions = grid.predict(test_data[['open', 'high', 'low', 'close', 'HL_PCT', 'PCT_change']])

Calculate mean absolute error to evaluate model performance
mae = mean_absolute_error(test_data['close'], predictions)
print("Mean Absolute Error: ", mae)

Plot actual vs predicted values
plt.figure(figsize=(12,8))
sns.lineplot(x=test_data.index, y=test_data['close'], label='Actual')
sns.lineplot(x=test_data.index, y=predictions, label='Predicted')
plt.legend()
plt.show()

Create function to enter