Long Short Term Memory for Forecasting the Stock Market

Abstract

The behavior of Stock market is similar to a random walk procedure and there is no relationship that exist in previous stock price and current stock price as it is statistically and economically ineffective. This is because of the impacts of undetermined exist in the market flow. In financial time-series forecasting, prediction of stock market is defined as demanding and arduous activity because of its highly dynamicnature, nonlinearities, cutoffs in the movement of other stock markets, macro economical parameters and many other political impacts. Therefore, it is necessary for predicting the stock market efficiently and effectively using forecasting tools to escalate the challenge of predicting the market flow. The accurate prediction model can be developed by analyzing the big data with the help of new technologies like data mining and machine learning.properties such as volatility, dependence and various related complex dependencies makes the model significant. Deep neural networks have recently been applied to forecast and predict the stock market. In our approach, Long Short Term Memory (LSTM) is a type of Recurrent Neural Network(RNN) is used for forecasting stock values for the Google Stock prices given.