Knowledge Patterns in Evolutionary Decision Support Systems for Financial Time Series Analysis
This paper discusses knowledge patterns in evolutionary learning of decision support systems for time series analysis, especially concerning time series of economical or financial data. It focuses on decision support systems, which use evolutionary algorithms to construct efficient expertises built on the basis of a set of specific expert rules analysing time series, such as artificial stock market financial experts composed of popular technical indicators analysing recent price quotations. Discovering common knowledge patterns in such artificial experts not only leads to an additional improvement of system efficiency, in particular - the efficiency of the evolutionary algorithms applied, but also reveals additional knowledge on phenomena under study. This paper shows a numer of experiments carried out on real data, discusses some examples of the knowledge patterns discovered in terms of their financial relevance as well as compares all the results with some popular benchmarks.
Unable to display preview. Download preview PDF.
- 1.Agrawal, R.A., Srikant, R.: Mining Sequential Patterns. In: Proc. of the International Conference on Data Engineering, ICDE, Taipei, Taiwan (1995)Google Scholar
- 2.Colby, W., Meyers, T.: The Encyclopedia of Technical Market Indicators, Down Jones-Irwin (1990)Google Scholar
- 5.Korczak, J., Lipinski, P.: Evolutionary Building of Stock Trading Experts in a Real-Time System. In: Proceedings of the 2004 Congress on Evolutionary Computation, CEC 2004, pp. 940–947. IEEE, Los Alamitos (2004)Google Scholar
- 7.Lipinski, P.: ECGA vs. BOA in Discoverying Stock Market Trading Experts. In: Proceedings of Genetic and Evolutionary Computation Conference, GECCO 2007, pp. 531–538. ACM, New York (2007)Google Scholar
- 10.Murphy, J.: Technical Analysis of the Financial Markets, NUIF (1998)Google Scholar