|
| | Abstract | | Natural earthquake tomography is widely used as a method of estimating subsurface structure. There is a least-squares method to one of them. But they depend on the initial condition. Therefore Oda and Ishiyama (2011) applied the neural networks used by various areas of the research to tomography. It is not dependent on the initial condition. However, since there was still a problem, it improved by adding the ray tracing using the Pseudo-Bending method. As a result, we can estimate the ground structure with good accuracy. |
|
|
|
| |
|