Some properties of adding a smoothing step to the EM algorithm

A promising method for solving statistical problems in image analysis and integral equations is to add a smoothing step after the usual expectation and maximization steps of the EM algorithm (Silverman, 1990). This article gives some connections between this algorithm, known as EMS, and maximizing a penalized likelihood and derives an upper bound on the convergence rate.Penalized likelihoods image analysis integral equations