Exclusive Content & Downloads from ASQ

Maximum Likelihood Estimation for Stochastic Differential Equations Using Sequential Gaussian-Process-Based Optimization

Summary: [This abstract is based on the authors' abstract.] Stochastic differential equations (SDEs) are used as statistical models in many disciplines. However, intractable likelihood functions for SDEs make inference challenging, and we need to resort to simulation-based techniques to estimate and maximize the likelihood function. While importance sampling methods have allowed for the accurate evaluation of likelihoods at fixed parameter values, there is still a question of how to find the maximum likelihood estimate. In this article, we propose an efficient Gaussian-process-based method for exploring the parameter space using estimates of the likelihood from an importance sampler. Our technique accounts for the inherent Monte Carlo variability of the estimated likelihood, and does not require knowledge of gradients. The procedure adds potential parameter values by maximizing the so-called expected improvement, leveraging the fact that the likelihood function is assumed to be smooth. Our simulations demonstrate that our method has significant computational and efficiency gains over existing grid- and gradient-based techniques. Our method is applied to the estimation of ocean circulation from Lagrangian drift data in the South Atlantic ocean.

Anyone with a subscription, including Site and Enterprise members, can access this article.

Other Ways to Access content:

Join ASQ

Join ASQ as a Full member. Enjoy all the ASQ member benefits including access to many online articles.

  • Topics: Engineering
  • Keywords: Maximum likelihood estimate (MLE), Simulation experiments, Stochastic models, Gaussian processes, Optimization, Differential equations
  • Author: Schneider, Grant; Craigmile, Peter F.; Herbei, Radu;
  • Journal: Technometrics