|
Activity Number:
|
110
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, August 7, 2006 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistics and the Environment
|
| Abstract - #306552 |
|
Title:
|
Estimation of Space-Time Branching Process Models in Seismology Using an EM-Type Algorithm
|
|
Author(s):
|
Alejandro Veen*+ and Frederic P. Schoenberg
|
|
Companies:
|
IBM T.J. Watson Research Center and University of California, Los Angeles
|
|
Address:
|
Statistical Analysis and Forecasting, Yorktown Heights, NY, 10598,
|
|
Keywords:
|
point process models ; branching process ; earthquakes ; seismology ; ETAS ; EM
|
|
Abstract:
|
The estimation of branching process models via a Maximum Likelihood can be unstable and computationally difficult. Viewing branching processes as incomplete data problems, however, suggests using the Expectation-Maximization algorithm as a practical estimation method. Using an application from seismology, we show that the Epidemic-type Aftershock Sequence (ETAS) model can in fact be estimated this way and we propose a particularly efficient procedure based on maximizing a partial log-likelihood function. Using a space-time ETAS model, we demonstrate that this method is extremely robust and accurate and use it to estimate declustered background seismicity rates of geologically distinct regions in Southern California.
|
