Design of an efficient Recurrent Neural Network based VARMA GRU Model for prediction of Inter-day stock movements via augmented indicator analysis

Traditional models for predicting stock value are limited in their ability to handle the complexity and dynamic nature of stock data, making the accurate prediction of intraday stock movements a significant challenge in the financial industry. This paper presents a Recurrent Neural Network (RNN)-based VARMA-GRU model to predict the fluctuations of intraday stock prices. The proposed model integrates an enhanced indicator analysis strategy into a VARMA-GRU model with technical indicators for the extraction of pertinent characteristics of stock data. The datasets & samples show nonlinear dependencies and long-term temporal dynamics, which are modeled by VARMA and GRU, respectively. The proposed model performs better than existing models in terms of accuracy, precision, recall, computational efficiency, and real-time practical use in stock market applications. The use case of this model is to support investors and traders in making informed decisions regarding the stock market by providing them with a very accurate prediction of intraday stock movements for different scenarios.