Improved Whale Optimization Algorithm with LSTM for Stock Index Prediction

After the COVID-19 pandemic, the global economy began to recover. However, stock market fluctuations continue to affect economic stability, making accurate predictions essential. This study proposes an Improved Whale Optimization Algorithm (IWOA) to optimize the parameters of the Long Short-Term Memory (LSTM) model, thereby enhancing stock index predictions. The IWOA improves upon the traditional Whale Optimization Algorithm (WOA) by integrating logistic chaotic mapping to increase population diversity and prevent premature convergence. Additionally, it incorporates a dynamic adjustment mechanism to balance global exploration and local exploitation, thus boosting optimization performance. Experiments conducted on five representative global stock indices demonstrate that the IWOA-LSTM model achieves higher accuracy and reliability compared to WOA-LSTM, LSTM, and RNN models. This highlights its value in predicting complex time-series data and supporting financial decision-making during economic recovery.