A simulated annealing version of the EM algorithm for non-Gaussian deconvolution

摘要: The Expectation–Maximization (EM) algorithm is a very popular technique for maximum likelihood estimation in incomplete data models. When the expectation step cannot be performed closed form, stochastic approximation of EM (SAEM) can used. Under general conditions, authors have shown that attractive stationary points SAEM correspond to global and local maxima observed likelihood. In order avoid convergence towards maxima, simulated annealing version proposed. An illustrative application convolution model estimating coefficients filter given.