An Analysis of the Performance of Artificial Neural Network Technique for Stock Market Forecasting

Abstract

In this study, we demonstrated a technique for daily stock price forecasting using neural networks, and we
compared the forecasting accuracy of the neural networks to that of statistical methods. Predicting stock prices using neural networks
is a relatively new topic of forecasting. Daily Stock Market Price Predictions using Neural Networks is also presented in this research.
Artificial neural networks are a common method for stock market forecasting, despite the fact that making accurate stock market
predictions is notoriously difficult due to the wide variety of known and unknown elements at play. The Neural Network uses the
'Learn by Example' principle as its foundation. This study models and forecasts daily stock market values using both Neural Networks
and statistical methodologies, and then compares the two sets of predictions. Using MAPE, MSE, and RMSE, we may evaluate the
accuracy of these two models for making predictions. The findings demonstrate that Neural Networks may accurately anticipate stock
market values after being trained with adequate data, appropriate inputs, and an appropriate design. Even while statistical methods are
robust, their ability to predict future series degrades as complexity increases. So, instead of relying on traditional methods, Neural
Networks may be utilized to make daily stock market price predictions