Stock prediction using artificial neural networks

• Main Author: Lim, Min
• Other Authors: PUN Chi Seng
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2021
• Subjects: Science::Mathematics; Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: https://hdl.handle.net/10356/146093
• Institution: Nanyang Technological University

Neural Networks are very good in learning patterns from a lot of information. Thus, it is generally used for character recognition, and can potentially be used for stock market prediction, if we assume the stock market actually follows a pattern (from certain information) and is not random. We start by obtaining S&P 500 stocks data QuantQuote Free Historical Stock Data [4]. We then construct the neural network and train the weights. For now, we have data of 500 stocks (from 1998 to 07/31/2013 for most of them). Our aim will be to use some previous days stock prices of a single stock to predict the next day stock prices (1 value to predict/output).