Stock Price Prediction Using LongShort Term Memory- Recurrent Neural Network Modelof ANN


  • D. Saveetha, K. Nimala, M. Sangeetha


The stock market exchange is known to be inconsistent and eccentric even with the plenitude of different AI and machine learning algorithms. These algorithms and calculations are utilized to anticipate the future estimation of a given stock over a significant period of time through data accessible to general society through web-sites and news. In spite of the fact that the forecast of the financial exchange has been an exemplary issue, there is no outright answer for it because the stock is resolved through several factors and its previous values are only a hint of something larger.In any case, there is an immense measure of studies from different fields trying to take on that challenge, introducing a huge assortment of ways to arrive at the objective.This paper goes towards that path however implementing a specific strategy utilizing recurrent neural networks. Such systems have a momentary memory capacity and the speculation to investigate here is that this component can introduce gains as far as results when contrasted with other increasingly conventional methodologies in the AI field. Long-Short-Term-Memory is a deep learning model that is utilized for time-series forecast and analysis. In this paper, we intend to foresee the estimations of a given stock by utilizing the past stock exchange information andthe LSTM model. This model will mechanize the way toward anticipating future stock value records, which will push different budgetary authorities and financial specialists to precisely purchase or sell stocks. The outcomes are shown in Python programming language. We have seen that the LSTM model has a real potential for finding an answer for financial exchange expectation.