Design and development of financial derivatives market prediction with neural networks

Abstract

Financial markets are widely available to the public and anyone can easily access and partake in it. It is a viable mean to generate passive income, resulting in investors constantly trying to find a better way to forecast prices. There are many available tools out there in the market, with strategies such as fundamental and technical analysis. Due to the recent development in computer technology, possible predictions could be made even more accurate with the help of neural networks. This project aims to research various types of neural network systems, and design an interface to showcase results. There are two parts to the project. First would be comparing between types of neural networks, to establish benchmark outcomes. Secondly, the predictive validity of this model is explored further with reference to various conditions. For example. there is sentimental analysis. Public sentiments would affect investors’ attitude towards companies, affecting changes in stock market. Anomaly detection would also be carried out, to warn investors should the market be at a volatile state. Hence, a multi-pronged approach would be proposed and developed to help investors make a more informed choice, yielding a higher rate of return.