475. A column generation approach to exact experimental design
Invited abstract in session WC-4: Large Scale Optimization for Statistical Learning, stream Optimization for machine learning.
Wednesday, 14:00-16:00Room: B100/5013
Authors (first author is the speaker)
| 1. | Selin Ahipasaoglu
|
| University of Southampton | |
| 2. | Stefano Cipolla
|
| School of Mathematical Sciences, University of Southampton | |
| 3. | Jacek Gondzio
|
| School of Mathematics, University of Edinburgh |
Abstract
We propose an algorithm for the exact optimal experimental design problem
under Kiefer’s criteria. Our method first employs column generation to solve the continuous relaxation of the problem quickly. The support of this solution is used to construct a feasible solution that is provably close to the optimal design. We demonstrate that for large-scale problems where the number of regression points is significantly larger than the number of experiments, this approach is preferable over existing algorithms.
Keywords
- Nonlinear mixed integer optimization
- Large-scale optimization
- Optimization for learning and data analysis
Status: accepted
Back to the list of papers