An Adaptive Recursive Least Square Algorithm for Feed Forward Neural Network and Its Application

Authors:
Xi-Hong Qing;Jun-Yi Xu;Fen-Hong Guo;Ai-Mu Feng;Wei Nin;Hua-Xue Tao
Affiliations:
College of Geo-Information Science and Engineering, Shandong, University of Science and Technology, 271019,Qingdao, Shandong, China;College of Geo-Information Science and Engineering, Shandong, University of Science and Technology, 271019,Qingdao, Shandong, China;College of Applied mathematics, Guangdong University of Technology, 510090,Guangzhou, Guangdong, China;Daqing Oilfield NO.2 Oil Production Company, 163414, Heilongjiang daqing, China;Shandong Agricultural University,271018,Taian Shandong, China;College of Geo-Information Science and Engineering, Shandong, University of Science and Technology, 271019,Qingdao, Shandong, China
Venue:
ICIC '07 Proceedings of the 3rd International Conference on Intelligent Computing: Advanced Intelligent Computing Theories and Applications. With Aspects of Artificial Intelligence
Year:
2009

Citing 4
Cited 0

Original Contribution: Optimal filtering algorithms for fast learning in feedforward neural networks

Neural Networks
A new Voronoi-based surface reconstruction algorithm

Proceedings of the 25th annual conference on Computer graphics and interactive techniques
2006 Special issue: Neural network forecasts of the tropical Pacific sea surface temperatures

Neural Networks - 2006 special issue: Earth sciences and environmental applications of computational intelligence
Developable surface modelling by neural network

Mathematical and Computer Modelling: An International Journal

Quantified Score

Hi-index	0.00

Visualization

Abstract

In high dimension data fitting, it is difficult task to insert new training samples and remove old-fashioned samples for feed forward neural network (FFNN). This paper, therefore, studies dynamical learning algorithms with adaptive recursive regression (AR) and presents an advanced adaptive recursive (AAR) least square algorithm. This algorithm can efficiently handle new samples inserting and old samples removing. This AAR algorithm is applied to train FFNN and makes FFNN be capable of simultaneously implementing three processes of new samples dynamical learning, old-fashioned samples removing and neural network (NN) synchronization computing. It efficiently solves the problem of dynamically training of FFNN. This FFNN algorithm is carried out to compute residual oil distribution.