Abstract
To solve a class of variational inequalities with separable structure, this paper presents a new method to improve the proximal alternating direction method (PADM) in the following senses: an iterate generated by the PADM is utilized to generate a descent direction; and an appropriate step size along this descent direction is identified. Hence, a descent-like method is developed. Convergence of the new method is proved under mild assumptions. Some numerical results demonstrate that the new method is efficient.
| Original language | English |
|---|---|
| Pages (from-to) | 17-29 |
| Number of pages | 13 |
| Journal | Computational Optimization and Applications |
| Volume | 49 |
| Issue number | 1 |
| DOIs | |
| Publication status | Published - May 2011 |
Scopus Subject Areas
- Control and Optimization
- Computational Mathematics
- Applied Mathematics
User-Defined Keywords
- Alternating direction method
- Descent method
- Proximal point method
- Separable structure
- Variational inequalities