My main research interest is nonlinear inference, especially within the context of dynamical systems, solved using probabilistic models and algorithms. In terms of scientific fields, my research is situated somewhere on the intersection between the fields of machine learning, signal processing and automatic control. My aim is to pursue both basic and applied research, where the latter is typically carried out in close collaboration with industry. A brief overview of my research is available here and my publications are available here.
Recent research results/news
August 22, 2016: Over the coming month Jack Umenberger is visiting us from Ian Manchester’s group at the University of Sydney for a short pre-doc. Together with Johan Wågberg we will work on designing EM-type algorithms for nonlinear system identification using recent Lagrangian relaxations and new particle smoothing solutions. Some early proof-of-concept work for linear systems is available here.
August 19, 2016: I am very glad to announce that Christian Andersson Naesseth will spend this academic year doing his pre-doc at Professor David Blei’s lab at Columbia University in New York. The work will be focused around the use of variational approximations in machine learning. Christian was awarded a Fulbright grant to finance his pre-doc.
August 18, 2016: We have developed a new class of sequential Monte Carlo (SMC) algorithms that is especially well suited for inference in probabilistic graphical models (including models with loops). By making use of an auxiliary tree-structured decomposition of the model we turn the original problem into a a collection of recursively solved sub-problems. This divide-and-conquer strategy has also given the name Divide-and-Conquer Sequential Monte Carlo (D&C-SMC) to the new class of algorithms. We illustrate the performance on a Markov Random Field (MRF) and on a hierarchical logistic regression problem. This work has now been accepted for publication in the Journal of Computational and Graphical Statistics (JCGS)
Fredrik Lindsten, Adam M. Johansen, Christian A. Naesseth, Bonnie Kirkpatrick, Thomas B. Schön, John Aston and Alexandre Bouchard-Côté. Divide-and-Conquer with Sequential Monte Carlo. Journal of Computational and Graphical Statistics (JCGS), 2016. [arXiv]
July 24, 2016: We have developed a Bayesian input design method for nonlinear state-space models. The key ingredients are Gaussian process optimization and the particle filter. This work will be presented at the 55th IEEE Conference on Decision and Control (CDC) that is held in Las Vegas in December.
Patricio E. Valenzuela, Johan Dahlin, Cristian R. Rojas and Thomas B. Schön. Particle-based Gaussian process optimization for input design in nonlinear dynamical models. In Proceedings of the 55th IEEE Conference on Decision and Control (CDC), Las Vegas, NV, USA, December, 2016. (accepted for publication) [arXiv]
July 5, 2016: Next week Pierre Jacob will present our coupling construction for the particle filter and the conditional particle filter at the World Congress in Probability and Statistics held in Toronto (Canada). All the details on the developments so far are available on arXiv.
Click here for older news.