Towards a Stock Price Prediction on Time Series Data using Long-Short Term Memory Method

The ongoing trend of stock market prediction using time series determines the financial trade exchange of a company. However, there are various challenges which include unexpected ups and downs, complex patterns, and sensitivity to radical changes. The exponential advancement in predictive analysis leads to determining the stock exchange on the basis of stored data. In this paper, deep learning techniques such as Long Short-Term Memory (LSTM) have been utilized for predicting the values. The LSTM method deals with long-term dependencies issues. Moreover, the forget gate along with the unit cell prevents LSTM from gradient vanishing and permits algorithms to store more than 1000 data points and process them smoothly. In experimental analysis, the results visualize the predictive values of stock with the actual stock values.