Pattern recognition applied to earthquake epicenters in California |
| |
Authors: | IM Gelfand ShA Guberman VI Keilis-Borok L Knopoff F Press EYa Ranzman IM Rotwain AM Sadovsky |
| |
Institution: | 1. Institute of Applied Mathematics, Academy of Sciences, Moscow U.S.S.R.;2. Institute of Physics of the Earth, Academy of Sciences, Moscow U.S.S.R.;3. Institute of Geophysics, University of California, Los Angeles, Calif. U.S.A.;4. Department of Earth and Planetary Sciences, Massachusetts Institute of Technology, Cambridge, Mass. U.S.A.;5. Institute of Geography, Academy of Sciences, Moscow U.S.S.R.;6. Laboratory of Mathematical Methods in Biology, Moscow University, Moscow U.S.S.R. |
| |
Abstract: | A pattern recognition procedure is explained which uses geological data and the earthquake history of a region, in this case California, and learns how to separate earthquake epicenters from other places. Sites of future earthquake epicenters are predicted as well as places where epicenters will not occur. The problem is formulated in several ways and control experiments are devised and applied in order to test the stability of the procedures and engender confidence in the results. Some of the combinations of geological features which the computer recognized as significant discriminants are discussed. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|