Stochastic search techniques are used when the structure of a space is not well understood or is not smooth, so that techniques like Newton’s method (which requires calculating Jacobian derivative matrices) cannot be used. In particular, these techniques are frequently used to solve combinatorial optimization problems, such as the traveling salesman problem.
The goal is to find a point in the space at which a real valued energy function (or cost function) is minimized. Simulated annealing is a minimization technique which has given good results in avoiding local minima; it is based on the idea of taking a random walk through the space at successively lower temperatures, where the probability of taking a step is given by a Boltzmann distribution.
The functions described in this chapter are declared in the header file gsl_siman.h.
|• Simulated Annealing algorithm:|
|• Simulated Annealing functions:|
|• Examples with Simulated Annealing:|
|• Simulated Annealing References and Further Reading:|