Studies of early comprehensive judgment on seismic tendency after a moderate or strong earthquake |
| |
| Authors: | Wei-Bin Han Hong Wang Jian Zeng and Dun-Li Xi |
| |
| Institution: | (1) Seismological Bureau of Sichuan Province, 610041 Chengdu, China |
| |
| Abstract: | In this paper, the methods of pattern recognition and fuzzy cluster are applied to estimate comprehensive whether there would
be an earthquake which magnitude is commensurate with (N=M
0±0.3) or greatM>M
0) thanM
0 within 3 months using sequential data in 3 days immediately after a moderate or strong earthquake (M
0≥4.7) occurred, The result of retrospective test indicates that the comprehensive judgment effects by using these applied
mathematics methods are better than that of using every individual method obviously. TheV mark value of algorithm CORA-3 in pattern recognition is the highest among them. There are 44 known earthquake sequence data
(5 type- I and 39 type- II) used to learn and train, the results of internal coincidence test show that all 44 sequence samples
could be distinguished correctly. Extrapolation test by using other 4 known earthquake sequences (2 type- I and 2 type- II)
shows that all 4 extrapolation samples could be distinguished correctly also. In the process of study, these methods have
been applied to judge the post-earthquake tendency of 2 moderate earthquakes occurred recently, one is distinguished correctly
and the other wrong. The algorithm Hamming in pattern recognition and fuzzy cluster method have been applied to judge the
early post-earthquake tendency after a moderate or strong earthquake, too. TheV mark values of internal coincidence tests could get above 0.8 most cases of extrapolating are correct.
The Chinese version of this paper appeared in the Chinese edition ofActa Seismologica Sinica,15, 15–21, 1993. |
| |
| Keywords: | post-earthquake tendency pattern recognition fuzzy cluster prediction mark |
| 本文献已被 SpringerLink 等数据库收录！ |
|
