|
Activity Number:
|
423
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 8:30 AM to 10:20 AM
|
|
Sponsor:
|
Section on Statistical Learning and Data Mining
|
| Abstract - #303781 |
|
Title:
|
Discriminant Analysis with Interval Constraints
|
|
Author(s):
|
Xiaoming Huo and Heeyoung Kim*+ and Jianjun Shi
|
|
Companies:
|
Georgia Tech and Georgia Tech and Georgia Tech
|
|
Address:
|
765 Ferst Drive, NW , Atlanta, GA, 30332,
|
|
Keywords:
|
discriminant analysis ; interval constraints ; convex optimization ; consistency ; classification
|
|
Abstract:
|
Efficient processing of high-dimensional data streams in manufacturing calls for discriminant analysis with interval constraints. We propose a two-stage algorithm, which is consistent, can be efficiently computed, and provides satisfactory results on real data sets. We refer to our method as discriminant analysis with interval constraints (DAIC) and demonstrate its application in tonnage data analysis for stamping processes.
|
