一种序贯学习神经网络及其应用
A SEQUENTIAL LEARNING ALGORITHM OF NEURAL NETWORK AND ITS APPLICATION
-
摘要: 本文提出了一种序贯学习神经网络,它主要由有界权值调整规则和网络结构自适应调整规则所组成。该网络具有在保持旧知识的前提下有效地序贯学习新输入样本知识的优点。文中给出了这种网络的一种序贯学习算法,详细分析了其学习特性和识别性能。大量的理论分析和实验都证明了网络的有效性。
-
关键词:
- 序贯神经网络; 学习算法; 模式分类
Abstract: In this paper we propose a sequential learning neural net which consists of the bounded weight adjustment algorithm and structure adaptive adjustment method. This network is characterized of efficiently learning the knowledge of new samples in series. After we present an efficient sequential learning algorithm of this network, we analyze its learning feature and recognition performance in detail. The effectiveness of this network have also been shown by theoretical analyses and a lot of experiments in this paper. -
Rao N S V, Protopopescu V, et al. Learning algorithms of feedforward networks based on finite samples[J].IEEE Trans. on Neural Networks.1996, 7(4):926-940[2]Hollis P W, Paulos J J. A neural network learning algorithm tailored for VLSI implementation[J].IEEE Tans. on Neural Networks.1994, 5(3):784-791[3]Rumelhart D E, Hinton G E, et al. Learning internal representation by error propagation. in Parallel Distributed Processing, Cambridge, MA: MIT Press, vol.l, 1986.[4]Carpenter G A, Grossberg S. ART2: Self-organization of stable category recognition codes for[5]analog input pattern. Proceedings of IEEE ICNN87, San Diego: 1987, 727-736.[6]Carpenter G A, Grossberg S, et al. Fuzzy ARTMAP: A neural-network architecture for incre-[7]mental supervised learning of analog multidimensional maps. IEEE Tans[J].on Neural Networks.1992, 3(3):698-713[8]Lim C P, Harrison R F. An incremental adaptive network for on-line supervised learning and probability estimation[J].Neural Networks.1997, 10(5):925-939[9]Lu Y W, Sundararajan N, Saratchandran P. A Sequential learning scheme for function approx-[10]imation using minimal radial basis function neural networks. Neural Computation, 1997, 9(2):[11]1-478.[12]Deng C.[J].Xiong F L, et al. Sequential learning neural net. Proceedings of 1998 IEEE International Joint Conference on Neural Networks (IJCNN98), Alaska, U.S.A: May.1328,1998:- -
计量
- 文章访问数: 2161
- HTML全文浏览量: 125
- PDF下载量: 661
- 被引次数: 0
下载:
