Neuro Fuzzy Systems for Data Analysis
Stefan Siekmann, Ralph Neuneier, Hans-Georg Zimmermann, Rudolf Kruse · Studies in fuzziness and soft computing · 1999
We present, how neural networks and fuzzy-systems can be combined to improve or create rules, which consist of linguistic expressions represented by fuzzy-sets. Due to the fuzzy-component we are able to integrate and extract expert knowledge. The neural component is used for optimization with historical data by transforming the rule base into a special neural network architecture. The parameters of these neural network are optimized with gradient descent techniques, which are combined with a semantic preserving algorithm. Therefore the optimized parameters can be transformed into an improved and still interpretable rule base. The special architectures enables us to change the form of the fuzzy-sets and the structure of the rule base. For structural optimization we use so called priming techniques on premises and rules. Rules can be deleted or changed by deletion and/or insertion of single premises. Also the creation of semantically correct rules is possible using this techniques. In section 1 and 2 an introduction to neuro-fuzzy-methods in financial data analysis is given. In chapter 3 different neural network architectures are presented, which can be applied for optimization of rule based systems. Section 4 shows, how to optimize a neuro-fuzzy-system using the proposed architectures and the learning algorithms of neural networks. To show the potential of the approach, we build neuro-fuzzy models for prediction of the daily returns of the German Stock Index DAX. The presented methods are implemented in the software environment for neural networks SENN of Siemens Nixdorf. A tutorial can be found at the web page http://www.sni-usa.com /snat/SENN/tutorial.