An Improved Minimum Distance Method Based on Artificial Neural Networks

Qing Li; Deling Zheng; Wenbo Meng; Yong Tang

Cite this article as:

Qing Li, Deling Zheng, Wenbo Meng, and Yong Tang, An Improved Minimum Distance Method Based on Artificial Neural Networks, J. Univ. Sci. Technol. Beijing , 9(2002), No. 1, pp.74-77.

引用本文 PDF XML SpringerLink

An Improved Minimum Distance Method Based on Artificial Neural Networks

摘要: MDM (minimum distance method) is a very popular algorithm in state recognition. But it has a presupposition, that is, the distance within one class must be shorter enough than the distance between classes. When this presupposition is not satisfied, the method is no longer valid. In order to overcome the shortcomings of MDM, an improved minimum distance method(IMDM) based on ANN (artificial neural networks) is presented. The simulation results demonstrate that IMDM has two advantages, that is, the rate of recognition is faster and the accuracy of recognition is higher compared with MDM.

An Improved Minimum Distance Method Based on Artificial Neural Networks

Abstract: MDM (minimum distance method) is a very popular algorithm in state recognition. But it has a presupposition, that is, the distance within one class must be shorter enough than the distance between classes. When this presupposition is not satisfied, the method is no longer valid. In order to overcome the shortcomings of MDM, an improved minimum distance method(IMDM) based on ANN (artificial neural networks) is presented. The simulation results demonstrate that IMDM has two advantages, that is, the rate of recognition is faster and the accuracy of recognition is higher compared with MDM.

HTML全文

资源附件(0)

施引文献