Title :
Neural classification approach for short term forecast of exchange rate movement with Point and Figure charts
Author :
Stahlbock, Robert
Author_Institution :
Inst. of Bus. Inf. Syst., Univ. of Hamburg, Hamburg
Abstract :
In the domain of classification and forecasting tasks, artificial neural networks (ANNs) are prominent data mining methods. Neural network paradigms like learning vector quantization (LVQ) are suitable for solving classification problems. In this paper, we combine LVQ with the popular Point & Figure (P&F) chart analysis applied to a one day forecast of the exchange rate between Euro (EUR) and US Dollar (USD). We present two different P&F encoding schemes and analyze the classification accuracy and results of a trading system fed with our results from the LVQ.
Keywords :
charts; data mining; economic forecasting; economic indicators; financial management; forecasting theory; learning (artificial intelligence); neural nets; pattern classification; vector quantisation; Euro-US Dollar exchange rate; artificial neural networks; chart analysis; classification problem; classification task; data mining method; exchange rate movement; figure charts; forecasting task; learning vector quantization; neural classification; neural network paradigm; point charts; short term forecast; trading system; Artificial neural networks; Chaos; Data mining; Economic forecasting; Exchange rates; Gaussian distribution; Globalization; International trade; Psychology; Vector quantization;
Conference_Titel :
Neural Networks, 2008. IJCNN 2008. (IEEE World Congress on Computational Intelligence). IEEE International Joint Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-1820-6
Electronic_ISBN :
1098-7576
DOI :
10.1109/IJCNN.2008.4634198
