| Autor: |
Dongming Wang, Lihong Zhi, Ming-Deh Huang, Qing Luo |
| Zdroj: |
Symbolic-Numeric Computation; 2007, p299-313, 15p |
| Abstrakt: |
The iterative proportional scaling algorithm is generalized to find real positive solutions to polynomial systems of the form: $$ \sum\nolimits_{j = 1}^m {a_{sj} p_j = c_s ,s = 1,...,n,} $$ where $$ p_j = \pi _j \prod\nolimits_{s = 1}^n {x_s^{a_{sj} } } $$ with asj ∈ ℝ and πj, cs ∈ ℝ>0. These systems arise in the study of reversible self-assembly systems and reversible chemical reaction networks. Geometric properties of the systems are explored to extend the iterative proportional scaling algorithm. They are also applied to improve the convergent rate of the iterative proportional scaling algorithm when dealing with ill-conditioned systems. Reduction to convex optimization is discussed. Computational results are also presented. [ABSTRACT FROM AUTHOR] |
| Databáze: |
Complementary Index |
| Externí odkaz: |
|
