对连续Hamming方法的改进及其在潜在震源判定中的应用a

AN IMPROVEMENT ON THE CONSECUTIVE HAMMING METHOD AND ITS APPLICATION TO DECIDING THE EARTHQUAKE-PRONE REGIONS

摘要: 本文提出了一种模式识别方法————改进的连续Hamming方法(ICHAM)。它从下面两个方面对使用连续资料的Hamming方法进行了改进。 1)考虑到更为一般的样品分布情况,用方差较小的那类样品计算Hamming核; 2)改进了对Hamming核的计算方法,使之处于方差较小的这一类样品的中心。文中还给出了ICHAM法应用于北京及邻区潜在震源判定的实例。结果表明,其效果比二值Hamming方法及原来未改进的连续Hamming方法都好一些。检验表明结果较为稳定。

Abstract: In this paper we propose a pattern recognition method, the improved consecutive Hamming method (ICHAM). It ameliorates the Hamming Method using consecutive data in two aspects: (1) considering the more general case, we compute the Hamming kernel using samples which belong to the class of small variance; (2) The algorithm computing the Hamming Kernel is improved so that it locates in the center of the class that the variance is small.An example, deciding the earthquake-prone regions of Beijing and its vicinity by the ICHAM method is shown in this paper. The results show that the effect of our ICHAM method is better than that of both the binary Hamming method and the original consecutive Hamming method. Tests show that the results are more stable.