Convergence rate of EM algorithm for SDEs under integrability condition

09/10/2020

∙

In this paper, by employing Gaussian type estimate of heat kernel, we establish Krylov's estimate and Khasminskill's estimate for EM algorithm. As applications, by taking Zvonkin's transformation into account, we investigate convergence rate of EM algorithm for a class of multidimensional SDEs under integrability conditions, where the drifts need not to be piecewise Lipschitz and are much more singular in some sense.

READ FULL TEXT

Convergence rate of EM algorithm for SDEs under integrability condition

On explicit L^2-convergence rate estimate for piecewise deterministic Markov processes

Singularity, Misspecification, and the Convergence Rate of EM

Accelerated and Deep Expectation Maximization for One-Bit MIMO-OFDM Detection

On the Semi-supervised Expectation Maximization

Explicit Numerical Approximations for SDDEs in Finite and Infinite Horizons using the Adaptive EM Method: Strong Convergence and Almost Sure Exponential Stability

The rate of Lp-convergence for the Euler-Maruyama method of the stochastic differential equations with Markovian switching

Residuum-Condition Diagram and Reduction of Over-Complete Endmember-Sets

Convergence rate of EM algorithm for SDEs under integrability condition

Related Research

On explicit L^2-convergence rate estimate for piecewise deterministic Markov processes

Singularity, Misspecification, and the Convergence Rate of EM

Accelerated and Deep Expectation Maximization for One-Bit MIMO-OFDM Detection

On the Semi-supervised Expectation Maximization

Explicit Numerical Approximations for SDDEs in Finite and Infinite Horizons using the Adaptive EM Method: Strong Convergence and Almost Sure Exponential Stability

The rate of Lp-convergence for the Euler-Maruyama method of the stochastic differential equations with Markovian switching

Residuum-Condition Diagram and Reduction of Over-Complete Endmember-Sets