A Novel Hybrid Model For Stock Price Forecasting: Combining Arima, Random Forests, And Gradient Boosting Techniques.

Accurately predicting stock prices is a complex challenge due to the volatile nature of financial markets. Traditional time series methods like ARIMA are effective in capturing linear trends but often struggle with complex, non-linear relationships. Recent advancements in machine learning, such as Random Forests and Gradient Boosting, offer improved modeling capabilities for intricate data patterns. This paper introduces a hybrid forecasting model that integrates ARIMA with Random Forests and Gradient Boosting to enhance stock price predictions. The approach starts by using ARIMA to model the linear components of stock price data, followed by the application of Random Forests and Gradient Boosting to the residuals to capture non-linear patterns. The performance of the hybrid model is assessed by generating and comparing prediction tables and plots for future stock prices. Results demonstrate that the hybrid model provides more accurate and reliable forecasts compared to the individual ARIMA, Random Forests, and Gradient Boosting models. This approach illustrates the potential of combining traditional time series analysis with advanced machine learning techniques to achieve superior stock price forecasting. [ABSTRACT FROM AUTHOR]