Machine Learning based Nifty 50 Index Forecasting Using Candlestick Pattern: A Systematic Analysis

The Nifty 50 index is an Indian stock market standard that measures the average value of the top 50 major country businesses registered on the National Stock Exchange (NSE). Nifty 50 index forecasting is a very difficult task during a highly volatile market, and this case-based study shows the approach to forecasting the Nifty 50 index value during a highly volatile market. This study is dependent on a combination of candlestick pattern-based and statistical-based technical indicators for the buying, selling opportunity, or investment purpose, a candlestick pattern can be analyzed with the help of deep learning models i.e. Convolution Neural network (CNN), Recurrent Neural Net- work(RNN), Long-short term Memory (LSTM) models along with different types of Keras applications while more than 20 application available. Statistical technical indicators help to find the momentum of the Nifty 50 index, and this momentum can be more accurate when applies technical indicators and candlestick pattern-based combinations. In this study, proposes a short-term (15-30 days) candlestick pattern-based strategy with different combinations of technical indicators and candlestick patterns. The graph is considered from Tradingview.com and the review mainly focuses on the Russia-Ukraine war. The Graph used from March 2022 onwards to November (2nd week) 2022 ended up with future impact. This study uses four indicators which are oscillator-based(RSI indicator), Trend following(MACD indicator), Volatility plotting(BB indicator), and instant trend forming(ADX indicator) with the combination of candlestick patterns. Finally, all four different indicators and candlestick patterns are combined and those strategies are briefly discussed.