Simulated Annealing: Parallelization TechniquesRobert Azencott This edited collection of papers, the result of an 18-month workshop held at France's Ecole Normale Superierure, covers the most current research on implementing the simulated algorithm on parallel computers. Professionals and students in mathematics, computer science, and electrical engineering will find theoretical results as well as actual simulations on existing parallel computers, sketching applications in discrete math, chip placement problems and spin glass models. |
Contents
Speed | 1 |
A Common Large Deviations Mathematical | 11 |
An Overview | 37 |
Copyright | |
7 other sections not shown
Common terms and phrases
Aarts acceleration acceptance rate accepted moves annealing schemes architecture Assume asymptotic configuration attraction domain average behavior Catoni Chapter communication computation configuration space constant cooling schedule cycle decreasing defined denote distribution elementary moves Emin energy function energy landscape evaluate experimental exponential f₂ Figure finite Freidlin Freidlin-Wentzell Geman given global minimum graph Henri Poincaré high temperature hypercube implementation inf H initial configuration Kirkpatrick local minima low temperature mode Markov chain matrix mesh method MIMD neighborhood neighbors number of iterations number of processors objective function observed probability obtained p₂ parallel algorithm parallel annealing parallel computer parallel simulated annealing parameter perturbations placement problem probability measure r₁ random rate of convergence Section sequence sequential algorithm sequential annealing SIMD simulated annealing algorithm solution speed of convergence spin glass SPMD synchronization T₁ temperature step tetrahedrons Theorem tion topology transputers updating