A Long Short-Term Memory Technique for NIFTY50 Stocks Price Prediction

One of the most debatable research areas has been stock price prediction which was not only a tough cookie to crack but also a real headache to work on. Despite the EMH assertions that stock prices cannot be rationally forecast as they represent and include all available information, empirical evidence has shown that informed modelling followed by appropriate variable selection is a lucrative milestone in highly accurate forecasting. However, the rise of artificial intelligence along with deep learning, new approaches to address this problem of complication will be presented. With the increasing complications of the future market due to the upsurge of complexity and uncertainty, artificial intelligence aspects especially Deep learning is the resolution offered. This study aims to analyze and forecast the closing price of the NIFTY 50 index by employing two advanced deep learning methods: LSTM and BE-LSTM. According to this 1 the type of data used in this study is from Bloomberg of 15 years of daily prediction which has a set of variables such as the last date close of data, high, open, low, close, volume, and the 14-period relative strength index. Comparative study reveals that BE-LSTM are innovative than the conventional LSTM in determining the NIFTY 50 index price over a day as compared to 30 days in which the prevalence accuracy it manages is 95% the findings estimate that BE-LSTM model Significanly enhances the accuracy of the NIFTY 50 closing price prediction.