|
Activity Number:
|
241
- Section on Statistics in Defense and National Security CPapers 1
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Monday, July 31, 2017 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistics in Defense and National Security
|
|
Abstract #323994
|
View Presentation
|
|
Title:
|
Interval Confidence Distributions for Tracking the North Korean 'Regression'
|
|
Author(s):
|
Robert Shumway*
|
|
Companies:
|
Univ Of California
|
|
Keywords:
|
Nuclear monitoring, yield estimation
|
|
Abstract:
|
Recent North Korean nuclear tests are alarming because they imply progressively larger explosions over time as measured by their estimated yields.This paper considers an initial calibration of yields to measured vectors of magnitudes using a multivariate regression model. This is followed by a subsequent inversion applied to a new observed magnitude vector to estimate the unknown yield. The problem falls naturally into the confidence distribution framework of Xie and Singh (2013) and we show that the predictive residuals can be employed to form a restricted two-sided confidence distribution involving Hotelling's T-Squared, expressed as a function of the unknown yield. We apply the methodology to four recent Korean tests and show graphical results in terms of "yield uncertainty profiles".
|
Authors who are presenting talks have a * after their name.
Back to the full JSM 2017 program
|
