Tools For Pattern Recognition In Univariate And Multivariate Processes
The objective of this research is to develop unified control charts that can be used to monitor and detect patterns in univariate and multivariate processes. These charts must becapable of locating shifts in, both process location and process variation, in a single chart. Currently, for the univaria...
Saved in:
| 主要作者: | |
|---|---|
| 格式: | Thesis |
| 出版: |
2007
|
| 主題: | |
| 標簽: |
添加標簽
沒有標簽, 成為第一個標記此記錄!
|
| 總結: | The objective of this research is to develop unified control charts that can be used to monitor and detect patterns in univariate and multivariate processes. These charts must becapable of locating shifts in, both process location and process variation, in a single chart. Currently, for the univariate case, the monitoring is done using two different charts, one for process mean and one for process variation. These charts are often cumbersome to construct and not easily interpreted by the quality control personnel. Individual quality characteristics are often measured in different scales of measures and thus having a standard chart is not possible. |
|---|