Analytical approximations for detection of a changepoint in case of light-tailed distributions
We derive analytic approximations for the expectation of exit times of Exponentially Weighted Moving Average (EWMA) procedure by using the martingale technique. Based on this technique, martingale approach is able to adapt to monitoring of changes of light-tailed distributions such as Gaussian, P...
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Article |
| Published: |
Penerbit ukm
2008
|
| Online Access: | http://journalarticle.ukm.my/1876/ http://www.ukm.my/~ppsmfst/jqma/index.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | We derive analytic approximations for the expectation of exit times of Exponentially
Weighted Moving Average (EWMA) procedure by using the martingale technique. Based on
this technique, martingale approach is able to adapt to monitoring of changes of light-tailed
distributions such as Gaussian, Poisson and Bernoulli distributions. Simple procedures are
addressed for obtaining the optimal design of EWMA. A comparison with Monte Carlo
simulation is also presented |
|---|