A Note on the Finite Time Behavior of Simulated Annealing

@article{Nolte2000ANO,
  title={A Note on the Finite Time Behavior of Simulated Annealing},
  author={Andreas Nolte and Rainer Schrader},
  journal={Math. Oper. Res.},
  year={2000},
  volume={25},
  pages={476-484}
}
Simulated Annealing has proven to be a very successful heuristic for various combinatorial optimization problems. It is a randomized algorithm that attempts to find the global optimum with high probability by local exchanges. In this paper we give a new proof of the convergence of Simulated Annealing by applying results about rapidly mixing Markov chains. With this proof technique it is possible to obtain better bounds for the finite time behavior of Simulated Annealing than previously known. 
Simulated Annealing and Graph Colouring
TLDR
It is shown for a certain class of graphs that the expected first hitting time of a proper colouring, given an arbitrary cooling scheme, is of exponential size.
IMPROVED BOUNDS FOR THE LARGE-TIME BEHAVIOUR OF SIMULATED ANNEALING
We improve on previous finite time estimates for the simulated annealing algorithm which were obtained from a Cheeger-like approach. Our approach is based on a Poincare inequality.
Analysis of Static Simulated Annealing Algorithms
TLDR
A measure for determining the expected number of iterations to visit a predetermined objective functionlevel, given that an inferior objective function level has been reached in a finite number of iteration, is introduced.
Simulated Annealing
TLDR
This chapter presents practical guidelines for the implementation of simulated annealing in terms of cooling schedules, neighborhood functions, and appropriate applications and establishes the connection between this type of thermodynamic behavior and the search for global minima for a discrete optimization problem.
Performance Analysis of Cyclical Simulated Annealing Algorithms
Generalized hill climbing (GHC) algorithms provide a framework for modeling local search algorithms for addressing intractable discrete optimization problems. Measures for assessing the finite-time
Dynamics on Landscapes 3 Simulated Annealing
TLDR
The presented algorithm is essentially the same as originally proposed by Kirkpatrick et al.
Stochastic Search Methods for Global Optimization
TLDR
This article highlights the use of a Markov chain Monte Carlo sampling method called Hit-and-Run in random search algorithms for global optimization.
A dynamic programming approach to efficient sampling from Boltzmann distributions
A simulated annealing algorithm for transient optimization in gas networks
TLDR
A simulated annealing approach for the gas network optimization problem where time steps as well as pressure and flow of the gas are decoupled and a suitable neighborhood structure is developed for the relaxed problem.
Evolutionary Optimization: Pitfalls and Booby Traps
TLDR
The goal is to equip practitioners and researchers alike with a clear picture and understanding of what kind of problems can render EC applications unsuccessful and how to avoid them from the start.
...
...

References

SHOWING 1-10 OF 27 REFERENCES
Convergence and finite-time behavior of simulated annealing
Simulated Annealing is a randomized algorithm which has been proposed for finding globally optimum least-cost configurations in large NP-complete problems with cost functions which may have many
Simulated Annealing and Its Problems to Color Graphs
TLDR
In this paper an application of Simulated Annealing to the 3-coloring problem is considered and it is shown that the expected first hitting time of a proper coloring, given an arbitrary cooling scheme, is of exponential size.
Markov Chains with Rare Transitions and Simulated Annealing
TLDR
A characterization of the convergent “cooling” schedules for the most general class of algorithms of the simulated annealing type is obtained.
Cooling Schedules for Optimal Annealing
A Monte Carlo optimization technique called “simulated annealing” is a descent algorithm modified by random ascent moves in order to escape local minima which are not global minima. The level of
Optimization by Simulated Annealing
TLDR
A detailed analogy with annealing in solids provides a framework for optimization of the properties of very large and complex systems.
Optimization by Simulated Annealing: An Experimental Evaluation; Part II, Graph Coloring and Number Partitioning
This is the second in a series of three papers that empirically examine the competitiveness of simulated annealing in certain well-studied domains of combinatorial optimization. Simulated annealing
Optimization by Simulated Annealing: An Experimental Evaluation; Part I, Graph Partitioning
TLDR
This paper discusses annealing and its parameterized generic implementation, describes how this generic algorithm was adapted to the graph partitioning problem, and reports how well it compared to standard algorithms like the Kernighan-Lin algorithm.
Simulated Annealing: Theory and Applications
TLDR
Performance of the simulated annealing algorithm and the relation with statistical physics and asymptotic convergence results are presented.
Large Cliques Elude the Metropolis Process
  • M. Jerrum
  • Computer Science
    Random Struct. Algorithms
  • 1992
TLDR
It is shown that the Metropolis process takes super-polynomial time to locate a clique that is only slightly bigger than that produced by the greedy heuristic, which is one step above the greedy one in its level of sophistication.
Ergodicity in Parametric Nonstationary Markov Chains: An Application to Simulated Annealing Methods
TLDR
The two ergodicity concepts are equivalent for finite chains under rather general (and widely verifiable) conditions and applications to probabilistic analyses of general search methods for combinatorial optimization problems (simulated annealing) are discussed.
...
...