Abstract
Control charts are well known tools capable to distinguish special and common causes of variation and to detect sustained changes in processes. However, these tools are not capable to estimate the initial moment of a sustained change. The detection and estimation of sustained changes belong to change point analysis. For series with shifts in variance, current models use control charts to detect the change and estimators based on different approaches to estimate the time of that change. The proposed approach uses the p-value function of the Chi-squared statistic to create a generalized likelihood ratio control chart capable of detecting shifts in variance and estimate the initial moment of a change at the same time.
Original language | American English |
---|---|
Pages | 2778-2782 |
Number of pages | 5 |
State | Published - 1 Jan 2012 |
Externally published | Yes |
Event | 62nd IIE Annual Conference and Expo 2012 - Duration: 1 Jan 2012 → … |
Conference
Conference | 62nd IIE Annual Conference and Expo 2012 |
---|---|
Period | 1/01/12 → … |
Keywords
- Change point
- Control chart
- Generalized likelihood ratio
- Statistical process control