Stock price prediction and forecasting using stacked LSTM in a smart environment

Abstract

The division that deals with money matters makes the most use of stock cost estimates. Predicting stock prices is challenging because of the inherent instability of the stock showcase. This is frequently a scheduling conflict. As there are no guidelines for estimating stock costs in the stock market, doing so can be difficult. There are currently many different ways to predict stock prices. Calculated Regression Model, SVM, Curve Show, RNN, CNN, Back Propagation, Naïve Bayes, ARIMA Demonstrate, etc. are examples of expectation strategies. Long short-term memory (LSTM) is the most logical model among them for time arrangement problems. Determining current advertising trends and accurately projecting stock costs are the main goals. We employ LSTM repetitive neural networks to accurately predict stock prices. The results seem to indicate that the predicted accuracy exceeds 93%.