Dynamic financial forecasting with automatically induced fuzzy associations
Yazann S. Romahi, Qiang Shen · 2002
The past decade has witnessed significant growth in developing intelligent tools for financial forecasting. Expert systems were quickly shown to be inadequate for the tasks required in financial forecasting due to their static nature. As a result, interest started to move towards soft computing despite the fact that comprehensibility is often of paramount concern in financial forecasting. Merging the domains of fuzzy logic and rule induction paved the way for the emergence of successful generalisation techniques with high comprehensibility. In this paper, we present a novel technique for financial forecasting derived from a fuzzy association induction algorithm, allowing the development of an evolving rule based expert system. In such a way, changing market dynamics are continuously taken into account as time progresses and thus the rulebase does not become outdated. Simulations carried out show promising results for this approach.