一种基于直推置信的遗传优化概率神经网络*

南京大学学报(自然科学版) ›› 2012, Vol. 48 ›› Issue (1): 48–54.

一种基于直推置信的遗传优化概率神经网络*

高志华¹，责可荣¹，章林柯^2**

出版日期:2015-05-21 发布日期:2015-05-21
作者简介: (1.海军工程大学计算机工程系，武汉，430033;
2.海军工程大学振动与噪声研究所，武汉，430033;
基金资助:
国家自然科学基金(50775218)

An optimized probabilistic neural network
based on transductive confidence

Gao Zhi一Hua¹，Ben Ke一Rong¹，Zhang Lin一Ke²

Online:2015-05-21 Published:2015-05-21
About author: (1 .Department of Computer Engineering, Naval University of Engineering, Wuhan, 430033，China
2, institute of Noise and Vibration, Naval University of Engineering, Wuhan, 430033，China)

摘要/Abstract

摘要： 提出一种基于直推置信机制的遗传优化概率神经网络用于水下航行器的机械噪声源识别.该网络模型采用遗传算法优化概率神经网络的结构，并且确定最优控制参数，在保证分类质量的同时提高
了分类器的识别速度.分类器在输出端引人直推置信机制，突破了传统分类器不能有效拒识突变样本的局限，通过置信机制实现无突变噪声训练样本情况下的突变噪声的识别.实验结果表明，该网络模型其
有良好的泛化性能、识别效果好，并且能有效识别突变噪声样本，是一种实用的水下航行器机械噪声源识别模型.

Abstract: An optimized probabilistic neural network for underwater vehicle mechanical noise source identification is proposed in this paper.The classifier used genetic algorithm optimizing neural network structure and optimal control
parameters, It improved the classifier recognition and ensured the quality of classification.The classifier can reject abnormity sample by confidence mechanism and realize training without noise samples.The experimental results
show the network model has good generalization and recognition performance. And it is a practical underwater vehicle mechanical noise source identification model

高志华¹，责可荣¹，章林柯^2**
. 一种基于直推置信的遗传优化概率神经网络*
[J]. 南京大学学报(自然科学版), 2012, 48(1): 48–54.

Gao Zhi一Hua¹，Ben Ke一Rong¹，Zhang Lin一Ke²
. An optimized probabilistic neural network
based on transductive confidence
[J]. Journal of Nanjing University(Natural Sciences), 2012, 48(1): 48–54.

参考文献

[1]Zhang I. K，He L, Zhu S J. Review on the methods of identification of submarine main noise sources. Noise and Vibration Control，2006, 26(4): 7-10.(章林柯，何琳，朱石坚. 潜艇主要噪声源识别为一法研究.噪声与振动控制，2006，26(4):7一10.
[2]Specht D F. Probabilistic neural networks. Neural Networks，1990，3(1)109一118.
[3]Simon L, Karim M N. Probabilistic neural net- works using Baycsian decision strategics and a modified Uompertz model for growth phase clas- sification in the batch culture of Bacillus subti- lis. Biochemical Engineering Journal，2001，7(1):41一18.
[4]Stecnhoek L W A. Probabilistic neural network computer vision system for corn kernel damage evaluation. PhD Thesis, Iowa State University, USA .1999.12-14
[5]Wang C J，Yu W D,Chen Z Q,et al. A BP neural network algorithm based on genetic algorithms and its application. Journal of Nanjing Univcrsity(Nat- ural Sciences),2003,39(5):459一466.(王崇骏，于
文将，陈兆乾等.一种基于遗传算法的YP神经网络算法及其应用.南京大学学报(自然科学)，2003.39(5).459 ~466)
[6]Zhao W B, Huang D S, Guo L. Genetic optimi- zation of the structure of radial basis probabilis- tic neural network. Journal of University of Sci- ence and Technology of China, 2003，33(6):
733一741.(赵温波，黄德双，郭磷.径向基概率神经网络结构的遗传优化.中国科技技术大学学报，2003, 33(6);733一741.
[7]Procdru K，Nourctdinov I，Vovk V, et al. Transductive confidence machine for pattern recognition. Elomaa T.Proceedings of the 13^th European Conference on Machine Learning. LNAI 2430，Heidelberg; Springer-Verlag, 2002，381一390.
[8]Barbara D, Domeniconi C，Rogers J P. Deter ting outliers using transduction and statistical testing. Ungar L, Craven M，Uunopulos D, et
al. Proceedings of the 12^th ACM SIUKDD In ternational Conference on Knowledge Discovery and Data Mining. New York; ACM Press,2006,55~64
[9]Vovk V，Gammerman A，Saunders C. Ma- chine-learning applications of algorithmic ran- dourness. Proceedings of the 16^th international Conference on Machine Learning, Bled，Sloveni- a,1999,444一453.
[10]Li Y，Fang Y X，Guo L, et al. A network anomaly detection method based on transduction scheme. Journal of Software, 2007，18(10): 2595-260.(李洋，为一滨兴，郭莉等.基于直推式为一法的网络异常检测为一法.软件学报，2007，18(10):2595一2604).
[11]Saunders C, Gammcrman A，Vovk V. Compu- tationally efficient transductive machines. Pro- ceedings of the 11th international Conference on Algorithmic Learning Theory, Stockholm, Sweden. 2000，325一337.

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed