Application of rough set theory to multi-factor medium and long-period runoff prediction in Danjing Kou reservoir

Authors:
Jing Guo;Wei Xiong;Hua Chen
Affiliations:
State Key Lab. of Water Resources and Hydropower Engineering Science, Wuhan University, Hubei, China;Radar and Avionics Institute of AVIC, Jiangsu, China;State Key Lab. of Water Resources and Hydropower Engineering Science, Wuhan University, Hubei, China
Venue:
FSKD'09 Proceedings of the 6th international conference on Fuzzy systems and knowledge discovery - Volume 5
Year:
2009

Citing 5
Cited 0

Rough Sets: Theoretical Aspects of Reasoning about Data

Rough Sets: Theoretical Aspects of Reasoning about Data
A Generalized Definition of Rough Approximations Based on Similarity

IEEE Transactions on Knowledge and Data Engineering
Bayesian framework for least-squares support vector machine classifiers, Gaussian processes, and kernel fisher discriminant analysis

Neural Computation
Neighborhood classifiers

Expert Systems with Applications: An International Journal
Fuzzy probabilistic approximation spaces and their information measures

IEEE Transactions on Fuzzy Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

The main objective of this study is to develop a predictor variable selection method based on rough set theory (RST) for runoff prediction, according to the different influence of different climate variables in different grid point on the runoff. The selected predictor variables were used as downscaling analysis predictors. Multiple linear regression (MLR), back propagation neural network (BPNN) and Bayesian least square support vector machine (Bay-LSSVM) statistical downscaling models were used to predict the monthly runoff of Danjiang Kou reservoir. NCEP/NCAR reanalysis data was utilized to establish the statistical relationship between the larger scale climatic predictors and observed runoff. Comparing with the performance of the statistical downscaling models without predictor variable selection, the models based on predictor variable selection were improved. Predictor variable selection based on RST not only reduced the dimension and noise of the predictor variables dataset greatly in each grid, but also enhanced the performance of statistical downscaling models.