Non-Convex Optimization
Harold H Szu · Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE · 1986
A stochastic search technique called simulated annealing can solve a class of problems termed non-convex optimization by seeking the lowest minimum of a multi-minima function. Simulated annealing is a generalized Monte Carlo technique with a continuously decreasing variance controlled by the temperature parameter.