Investigating the Impact of Data Volume on Stock Market Prediction: Insights from Artificial Neural Networks

Abstract

This study investigates the impact of varying data quantities on stock market prediction using Artificial Neural Networks (ANNs). It assesses short-term and long-term forecasts, considering time delays. ANNs, inspired by the human brain, are adept at recognizing patterns in intricate systems, making them suitable for stock market prediction. The research focuses on data from major IT and Telecom companies in the OMX30 Stockholm index. It utilizes specialized networks trained via supervised learning, conducting thorough testing to identify optimal setups. Results suggest that for short-term forecasts, reduced time delays lead to improved accuracy, and optimal configurations remain consistent with increasing data volume. However, conclusive insights for long-term predictions are not provided.