Hybrid Approach Using LSTM Neural Networks and Whale Optimization Algorithm For Enhancing Depression Diagnosis

Abstract

Depression is a prevalent mental disorder, and accurate diagnosis is essential for effective intervention by proposing a novel approach for diagnosing depression using Long Short-Term Memory (LSTM) neural networks combined with swarm intelligence algorithms. Accordingly, this research pays major attention to how the strength of LSTM in sequential data and optimization performance through swarm intelligence enhance the accuracy and reliability of the diagnosis of depression. So far, we utilized the Whale Optimization Algorithm (WOA) algorithm for the optimal tuning of the model parameters of LSTM toward their optimal performance. The experimental results confirm the effectiveness of our approach, where the proposed model has an RMSE value of 3.75 and an MAE value of 3.15. These results indicate a higher performance compared to state-of-the-art methods and offer great potential for LSTMs combined with the WOA algorithm in the accurate diagnosis of depression. This may be a signal that the hybrid approach will serve perfectly in a clinical setting by allowing clinicians to identify depression among patients more precisely and reliably.