Dynamic Stock Price Prediction Leveraging LSTM, ARIMA, and Sparrow Search Algorithm

Abstract

This research paper presents a novel approach to stock price estimation utilizing a hybrid model that combines the strengths of Long Short-Term Memory (LSTM) and the Autoregressive Integrated Moving Average (ARIMA) models. The LSTM model is renowned for capturing long-term dependencies and intricate patterns in sequential data, while ARIMA provides a robust statistical framework for time series forecasting, adept at capturing short-term trends and seasonality. By leveraging these complementary strengths, the aim is to enhance predictive accuracy across various stock market conditions.

To optimize the model's performance, the Sparrow Search Algorithm (SSA), inspired by the foraging behavior of sparrows, is introduced. This algorithm efficiently explores the hyper parameter space to identify the optimal configuration for the model. By dynamically adjusting parameters such as learning rates, batch sizes, and network architectures, the SSA ensures superior performance and adaptability of the hybrid model. 

Through extensive experimentation with historical stock market data, the efficacy of the proposed approach is evaluated. The model undergoes rigorous testing, including back testing and validation across multiple stocks and market scenarios to assess its accuracy and robustness. Results demonstrate significant improvements in predictive performance compared to other models, highlighting the effectiveness of the hybrid approach and SSA in enhancing stock price estimation techniques.