Author(s): C. D. Liong S. Y. Doan; D. S. K. Karunasingha; C. H. Ong
Linked Author(s):
Keywords: Fuzzy Inference System; Neural Networks; Subtractive Clustering Method; Genetic Algorithm; Chaos
Abstract: Success of any forecasting model depends heavily on reliable historical data. Data are needed to calibrate, fine tune, and verify any simulation model. However, data are very often contaminated with noises of different levels originated from different sources. This study proposed a scheme that extracts mainly most representative data from raw data set. Subtractive Clustering Method (SCM) and Micro Genetic Algorithm (mGA) were used for this purpose. SCM does (a) remove outliers; and (b) discard unnecessary or superfluous points. mGA, a search engine, determines the optimal values of the SCM's parameter set. The scheme was demonstrated in (1) Bangladesh water level forecasting with Neural Network and Fuzzy Logic; and (2) discharge forecasting of Wabash River at Mt. Carmel and Mississippi River at Vicksburg with Chaos Theory. The scheme significantly reduces the data set with which the forecasting models yield either equally high or higher prediction accuracy than models trained with the whole original data set. The resulting fuzzy logic model, for example, yields a less number of rules which are easier for human interpretation. In chaos analysis, which is known to require long data record, a data reduction of up to 60% affects the prediction accuracy only by about 10%.
Year: 2003