A Hybrid Attention-BiGRU Model for Accurate Stock Market Prediction

In today's fast-paced financial markets, investors are constantly seeking strategies to gain an edge and make informed decisions. While achieving 100 % accuracy in stock price predictions remains a challenge, advancements in artificial intelligence (AI) have made it possible to examine past data and uncover potential trends. This article explores the application of AI techniques, specifically the Bidirectional Gated Recurrent Unit (BiGRU) combined with an Attention Mechanism, for stock price prediction. These tools enable investors to better understand market behaviour and make well-informed investment decisions. In this study, datasets from CL=F, DAX, <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">${ }^{\wedge}$</tex> GDAXI, and <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">${ }^{\wedge}$</tex> HSI were utilized. The results indicated that the proposed model successfully achieved RMSE and MAE values ranging between 0.011 to 0.065 across all stock datasets.