STUDY ON THE METEOROLOGICAL PREDICTION MODEL USING THE LEARNING ALGORITHM OF NEURAL ENSEMBLE BASED ON PSO ALGORITHMS

WU Jian-sheng; JIN Long

Abstract:

Because of the difficulty in deciding on the structure of BP neural network in operational meteorological application and the tendency for the network to transform to an issue of local solution, a hybrid Particle Swarm Optimization Algorithm based on Artificial Neural Network (PSO-BP) model is proposed for monthly mean rainfall of the whole area of Guangxi. It combines Particle Swarm Optimization (PSO) with BP, that is, the number of hidden nodes and connection weights are optimized by the implementation of PSO operation. The method produces a better network architecture and initial connection weights, trains the traditional backward propagation again by training samples. The ensemble strategy is carried out for the linear programming to calculate the best weights based on the “east sum of the error absolute value” as the optimal rule. The weighted coefficient of each ensemble individual is obtained. The results show that the method can effectively improve learning and generalization ability of the neural network.

Get Citation+

WU Jian-sheng, JIN Long. STUDY ON THE METEOROLOGICAL PREDICTION MODEL USING THE LEARNING ALGORITHM OF NEURAL ENSEMBLE BASED ON PSO ALGORITHMS [J]. Journal of Tropical Meteorology, 2009, 15(1): 83-88.

Export:

Article Metrics

STUDY ON THE METEOROLOGICAL PREDICTION MODEL USING THE LEARNING ALGORITHM OF NEURAL ENSEMBLE BASED ON PSO ALGORITHMS

Department of Mathematic and Computer Sciences, Liuzhou Teachers College

Abstract: Because of the difficulty in deciding on the structure of BP neural network in operational meteorological application and the tendency for the network to transform to an issue of local solution, a hybrid Particle Swarm Optimization Algorithm based on Artificial Neural Network (PSO-BP) model is proposed for monthly mean rainfall of the whole area of Guangxi. It combines Particle Swarm Optimization (PSO) with BP, that is, the number of hidden nodes and connection weights are optimized by the implementation of PSO operation. The method produces a better network architecture and initial connection weights, trains the traditional backward propagation again by training samples. The ensemble strategy is carried out for the linear programming to calculate the best weights based on the “east sum of the error absolute value” as the optimal rule. The weighted coefficient of each ensemble individual is obtained. The results show that the method can effectively improve learning and generalization ability of the neural network.

HTML

Reference (20)

[1]	HU Jian-lin, TU Song-bo, FENG Guang-liu. Anexploration of heavy rain forecasting technique based onartificial neural networks [J]. J. Trop. Meteor., 2003, 19(4):422-428.
[2]	HSIEH W W. Nonlinear canonical correlation analysis ofthe tropical Pacific climate variability using Neural NetworkApproach [J]. J. Climate, 2001, 14(12): 2528-2539.
[3]	GIORGIO C, GIOGIO G. Coupling Fuzzy Modeling andNeural Networks for River Flood Prediction [J]. IEEETransactions on Systems, Man, and Cybernetic-Part C:Applications and Reviews, 2005, 25(3): 382-388.
[4]	WU Jian-sheng, JIN Long, WANG Ling-zhi. The backpropagation neural network meteorological forecast modelresearch evolved and designed by genetic algorithms [J]. J.Trop. Meteor., 2006, 22(4): 411-416.
[5]	HE Hui, JIN Long, QING Zhi-nian, et al. Downscalingforecast for the monthly precipitation over Guangxi based onthe BP neural network model [J]. J. Trop. Meteor., 2007, 23(1):72-77.
[6]	JIN Long, KUANG Xue-yuan, et al. Study on theover-fitting of the artificial neural network forecasting model[J]. Acta Meteor. Sinica, 2004, 62(1): 62-69.
[7]	HANSEN L K, SALAMON P. Neural network ensembles[J]. IEEE Transactions on Pattern Analysis and MachineIntelligence, 1990, 12(10): 993-1001.
[8]	SOLLICH P, KROGH A. Learning with Ensembles: HowOver-fitting can be useful [C]// Advances in NeuralInformation Processing Systems 8, Cambridge: MIT Press,1996: 190-196.
[9]	ZHOU Zhi-hua, CHEN Shi-fu. Neural network ensemble[J]. Chin. J. Comput., 2002, 25(1): 1-8.
[10]	MAO J. A case study on bagging boosting and basicensembles of neural networks for OCR [C]// ProcessingInternational Joint Conference on Neural Networks 1998.Anchorage: International Joint Conference on NeuralNetworks, 1998: 1828-1833.
[11]	GUTTA S, WECHSLER H. Face recognition using hybridclassifier systems [C]// Proceeding International JointConference on Neural Networks 1996. Washington DC: 1996:1017-1022.
[12]	SOLLICH P, INTRATOR N. Classification of seismicsignals by integrating ensembles of neural networks [J]. IEEETrans. Signal Process., 1998, 46(5): 1194-1021.
[13]	LI NING, ZHOU HUA-JIE, LING JIN-JIANG, et al.Speculated lesion detection in digital mammogram based onartificial neural network ensemble [J]. Adv. Neural NetworksISNN, Springer Press, 2005, 3: 790-795.
[14]	BONABEAU E, DORIGO M, THERAULAZ G.Inspiration for optimization from social insect behavior [J].Nature, 2000, 406(6): 39-42.
[15]	XIAOHUI H, EBERHART R. Multi-objectiveoptimization using dynamic neighborhood particle swarmoptimization [C]// Proceeding of Congress on EvolutionaryComputation. Hawaii: Congress on Evolutionary Computation,2002: 1677-1681.
[16]	RIGET J, VESTERSTROM J S. A diversity-guidedparticle swarm optimizer-the ARPSO [R]. Technical Report2002-02, Department of Computer Science, University ofAarhus, 2002.
[17]	MA Zhen-hua. Operations Research and OptimizingTheory [M]. Beijing: Tsinghua Press. 1998, 235-425.
[18]	VAUTARD. SSA: a toolkit for noisy chaotic signals [J].Physical D, 1992, 58: 95-126.
[19]	WEI Feng-ying, CHAO Hong-xing. The MathematicsForecast Model and Application of Long Period Time [M].Beijing:Meteorological Press, 1990.
[20]	WANG Hui-weng. The Model and Application of PartialLeast-Squares Regression [M]. Beijing: National DefensesScience and Technology University Press, 1999, 258-365.

STUDY ON THE METEOROLOGICAL PREDICTION MODEL USING THE LEARNING ALGORITHM OF NEURAL ENSEMBLE BASED ON PSO ALGORITHMS

Abstract:

References

Get Citation+

Share Article

Article Metrics

Proportional views

Manuscript History

Online:

通讯作者: 陈斌, bchen63@163.com