Golden Eagle optimized Hybrid RNN-GRU Model for Stock Price Prediction: A Data-Driven Deep Learning Approach

Predicting stock prices accurately remains a formidable challenge due to their volatile and non-linear behavior. Advent of Artificial Intelligence (AI) and Machine Learning (ML) significantly enhanced computational capabilities, improving efficiency in stock price forecasting. This research employs Deep Learning (DL) methods to predict future stock prices, leveraging a Hybrid Recurrent Neural Network (RNN)-Gated Recurrent Unit (GRU) integration to boost prediction accuracy. Hybrid RNN-GRU model effectively captures the complex patterns in stock price deviations across diverse market sectors. The performance of hybrid RNN-GRU model is further enhanced by finely tuning its hyperparameters by optimization algorithms. Golden Eagle Optimization (GEO) approach is utilized which mimics hunting behavior of eagle to explore and exploit optimal solution effectively in predicting stock price trends. Evaluation of the proposed model includes performance metrics such as Mean Square Error (MSE) at 0.1045, Root Mean Square Error (RMSE) at 0.4189, Mean Absolute Error (MAE) at 0.9985, and an R² score of 0.9980. These relatively low error values demonstrate model's high efficiency in predicting stock prices accurately.