A study on stock price prediction using LSTM Model

Abstract

Stock market is an industry where lots of data is generated daily and benefits are reaped on the basis of accurate prediction. Many people invest in stock market having some prediction and more luck. A decision in stock market plays an important role in the investor’s life. Also, stock market is a very complex system and non-linear in nature. So, then it is very difficult to analyse all the impacting factors before making a decision. Making decision with traditional techniques may be time consuming and may not ensure the reliability of the prediction. Data from stock market is a time series data and different variations of neural networks are widely being used for stock forecasting and prediction problems. Among various architectures of deep neural networks, Long Short-Term Memory (LSTM) is one that supports time steps of arbitrary sizes and is free of vanishing gradient problem. Furthermore, stock price analysis is becoming popular in predicting the stock market. Research in recent years has shown that LSTM) network models perform well in stock price prediction, and it is considered one of the most precise prediction techniques, especially when it is applied to longer prediction ranges. This research attempts to explore the capabilities of Long Short-Term Memory a type of Recurrent Neural Networks in the prediction of future stock prices. We will use multiple layers of neurons to train the model and try to predict the actual price using different dataset of the stocks for the upcoming days. It is assumed that the prediction will be done for the shorter timeframe.