Stock Market Prediction Using RNN LSTM

Abstract

Bond price prediction is a trendy, demanding, hard and complicated problem in the realm of computation that usually includes considerable interaction between people and computers. The trends for predicting stock market physical aspects versus physiological, rational, and illogical conduct, investor emotion, market whispers are engaged in various factors. All those facets mix to make stock values extremely sophisticated and exceedingly difficult to accurately anticipate. [15] Because of the linked nature of stock prices, sequential prediction algorithms may be used for stock market prediction efficiently. ML methods can identify patterns and determine the logic and forecasts and can be utilized to produce unerringly correct predictions. [8] We have explored several different algorithms to forecast the stock market from simple algorithms, such as Simple Average, Linear regression, to advance algorithms like ARIMA, LSTM, and compare what gives us a more accurate result and works more efficiently. We offer a research technology that employs the improved Long Short Term Memory (LSTM) version of RNN, with stochastic gradient descent maintaining the weights for each data variable. [8] To help us deliver more efficient and accurate outcomes than existing stock price prediction systems. We have utilized the TSLA dataset to create the stock prediction model: this is TESLA Inc. from Yahoo Finance. We have analyzed future stock prices using data-frame closing prices, built up and trained the LSTM model, and have taken a data set sample to generate stock forecasts and computed additional RMSE for correctness and effectiveness. We have also displayed several algorithms for comparative predictions, Based on these outcomes, LSTM is recommended for stock market forecasts.