Randomized Algorithms for Semi-Infinite Programming Problems [chapter]

Vladislav B. Tadić, Sean P. Meyn, Roberto Tempo
Probabilistic and Randomized Methods for Design under Uncertainty  
This paper studies the development of Monte Carlo methods to solve semi-infinite, nonlinear programming problems. An equivalent stochastic optimization problem is proposed, which leads to a class of randomized algorithms based on stochastic approximation. The main results of the paper show that almost sure convergence can be established under relatively mild conditions.
doi:10.1007/1-84628-095-8_9 fatcat:kii7y3trdffabivlmrkcm62zvu