An Efficient Approach to Forecasting the NIFTY-50 Indian Stock Market's Daily Closing Price with Artificial Neural Networks

Abstract

The most lucrative area of the financial industry is stock market trend forecasting. The incorporation of machine learning has improved the accuracy, accessibility, and dependability of stock market forecasts. This study compares and contrasts several machine learning methods, including Support Vector Machine (SVM), Random Base Pairing (RBF), Multi-Layer Perceptron (MLP), and Single Layer Perceptron (SLP). The unique features of each method are also examined separately. Of these methods, the MLP algorithm produced better results. The inquiry explores the relationship between changes in the dollar exchange rate, the performance of NIFTY 50, Foreign Institution Investors' Gross Purchase (FII inflow), and Foreign Institution Investors' Gross Sale (FII outflow). For the study, daily average closing prices for the NIFTY 50, FII inflow, FII outflow, and Dollar are taken into account. The study covers a period of 5 years from April 1, 2018, to March 31, 2023. The NN Toolbox in Matlab is used in the analysis to evaluate the correlation between NIFTY 50, FII inflow, FII outflow, and Dollar exchange rates. A strong correlation has been found by the data analysis between Dollar values, FII inflow, FII outflow, and NIFTY 50.