Saved in:
| Main Author: | |
|---|---|
| Format: | Recurso digital |
| Language: | |
| Published: |
Zenodo
2026
|
| Online Access: | https://doi.org/10.5281/zenodo.18765378 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Table of Contents:
- <p><span>This paper proposes a novel optimization theoretical framework—Information Polarized Conic Programming (IPCP). This framework establishes a rigorous geometric-information unified structure between general cone programming and φ-divergence DRO (Divergence-Related Robust Optimization). By constructing the "information polarized cone" and proving its closed convex cone property, we embed distributional uncertainty into an extended cone space, achieving a cone-shaped expression of the probability measure uncertainty set. This paper establishes the information polarization duality theory, strong duality theorem, polarization stability theorem, and structure collapse theorem, proving that the model possesses a unified dual structure at both the geometric and informational levels. The IPCP framework rigorously includes classical cone programming and φ-divergence DRO as special cases, and provides new geometric tools for robust optimization and statistical learning theory</span>.</p>