
Identification of Gaussian Process StateSpace Models with Particle Stochastic Approximation EM
Gaussian process statespace models (GPSSMs) are a very flexible family...
read it

Convergence of the ExpectationMaximization Algorithm Through DiscreteTime Lyapunov Stability Theory
In this paper, we propose a dynamical systems perspective of the Expecta...
read it

A heuristic independent particle approximation to determinantal point processes
A determinantal point process is a stochastic point process that is comm...
read it

The EM algorithm and the Laplace Approximation
The Laplace approximation calls for the computation of second derivative...
read it

Maximum likelihood estimation of potential energy in interacting particle systems from singletrajectory data
This paper concerns the parameter estimation problem for the quadratic p...
read it

A Dynamical Systems Approach for Convergence of the Bayesian EM Algorithm
Out of the recent advances in systems and control (S&C)based analysis o...
read it

MLEM algorithm with known continuous movement model
In Positron Emission Tomography, movement leads to blurry reconstruction...
read it
Learning dynamical systems with particle stochastic approximation EM
We present the particle stochastic approximation EM (PSAEM) algorithm for learning of dynamical systems. The method builds on the EM algorithm, an iterative procedure for maximum likelihood inference in latent variable models. By combining stochastic approximation EM and particle Gibbs with ancestor sampling (PGAS), PSAEM obtains superior computational performance and convergence properties compared to plain particlesmoothingbased approximations of the EM algorithm. PSAEM can be used for plain maximum likelihood inference as well as for empirical Bayes learning of hyperparameters. Specifically, the latter point means that existing PGAS implementations easily can be extended with PSAEM to estimate hyperparameters at almost no extra computational cost. We discuss the convergence properties of the algorithm, and demonstrate it on several machine learning applications.
READ FULL TEXT
Comments
There are no comments yet.