|
Activity Number:
|
67
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Sunday, August 6, 2006 : 4:00 PM to 5:50 PM
|
|
Sponsor:
|
Section on Quality and Productivity
|
| Abstract - #306975 |
|
Title:
|
Hierarchcial Modeling Using GLMs To Improve Yield
|
|
Author(s):
|
Christina Mastrangelo*+ and Naveen Kumar
|
|
Companies:
|
University of Washington and University of Washington
|
|
Address:
|
2319 44th Ave., SW, Seattle, WA, 98116,
|
|
Keywords:
|
generalized linear models ; hierarchical modeling
|
|
Abstract:
|
In a complex manufacturing environment such as semiconductor manufacturing, there are hundreds of interrelated processes. In such an environment, modeling the impact of critical process parameters on final performance metrics such as defectivity or yield is a challenging task. Issues such as low number of observations compared to process variables, difficulty in formulating a high dimensional design matrix, and missing data due to failures pose serious challenges in using empirical modeling techniques. Our approach is to use generalized linear modeling in a hierarchy to understand the impact of key process and subprocess variables on the system output. Issues such as bias and variance estimation are considered. The hierarchical GLM approach helps not only in improving output metrics, but also in identifying and improving subprocess variables attributable to poor performance.
|
