Exploring The Synergy: Combining Improved Sparrow Search Optimization with Long Short-Term Memory (LSTM) For Enhanced Stock Price Prediction

Abstract

The present study employs an improved Sparrow Search Optimization (SSA) technique to optimize hyperparameters in Long Short-Term Memory (LSTM) networks to predict stock prices. The present study adopts an innovative approach that leverages the augmented SSA's advanced capabilities to optimize the LSTM model's hyperparameters and enhance prediction accuracy. The evaluation measures are employed to test the performance of the proposed SSA-LSTM model on a specific stock market dataset. Our findings demonstrate that the SSA-LSTM model surpasses a conventional LSTM model when employing traditional tuning of hyperparameter methods. This gain showcases the improved synergy between the augmented SSA and the LSTM model, suggesting it can provide more precise stock price predictions.