State Space Emulation and Annealed Sequential Monte Carlo for High Dimensional Optimization

Many high dimensional optimization problems can be reformulated into a problem of finding theoptimal state path under an equivalent space model setting. In this article, we present general emulation strategy for developing whose likelihood function (or posterior distribution) shares the same landscape as objective original problem. Then solution is optimal that maximizes or distribution emulated system. To find such path, adapt simulated annealing approach by inserting temperature control dynamic system and propose novel annealed Sequential Monte Carlo (SMC) method effectively generating sample paths utilizing samples obtained on scale. Compared to vanilla implementation, SMC iterative algorithm directly generates from equilibrium distributions with decreasing sequence temperatures through sequential importance sampling which does not require burn-in mixing iterations ensure quasi-equilibrium condition. Several applications corresponding results are demonstrated.