We present a randomized algorithm called Model Reference Adaptive Search (MRAS) for solving global optimization problems. The algorithm generates at each iteration a group \nof candidate solutions according to a parameterized probabilistic model. These candidate solutions are then used to update the parameters associated with the probabilistic model in such a way that the future search will be biased toward the region containing high quality solutions. The parameter updating procedure in MRAS is guided by a sequence of implicit reference models that will eventually converge to a model producing only the optimal\nsolutions. We establish global convergence of MRAS in both continuous and combinatorial domains.\nNumerical studies are also carried out to demonstrate the effectiveness of the algorithm. | 
