IRGM: An Integrated RNN-GRU Model for Stock Market Price Prediction

Abstract

Stock market price prediction has been very challenging because of the non-linearity of such a dataset. Several recent research studies have demonstrated the value of accurate prediction as a motivator for investors, researchers, and market analysts, regardless of market trend. Traditional approaches such as Linear Regression (LR) and Support Vector Machines (SVM) have produced results that are inefficient and inaccurate, leading to the development of Deep Learning-based solutions (DL). With the increased use of DL techniques in prediction, its performance is showing significant improvements over traditional time series models. With their ability to learn patterns and predict future prices, DL models such as Convolutional Neural Networks (CNN), and Recurrent Neural Networks (RNN) are becoming increasingly popular. Even so, these models continue to evolve and are continually revised to make them more efficient and better. To forecast the price for the 11th day, we present an RNN-GRU (Gated Recurrent Unit) model based on four independent features and the prices from the previous ten days. Furthermore, this model's performance was compared to that of an LSTM and CNN model. The results of the last test revealed that our proposed model outperformed the CNN and LSTM models.