Multi-view Matrix Factorization for Linear Dynamical System Estimation
- Mahdi Karami
- Martha White, University of Alberta
- Dale Schuurmans, AICML
- Csaba Szepesvari, Department of Computing Science; PI of AICML
We consider maximum likelihood estimation of linear dynamical systems with generalized-linear observation models. Maximum likelihood is typically considered to be hard in this setting since latent states and transition parameters must be inferred jointly. Given that expectation-maximization does not scale and is prone to local minima, moment-matching approaches from the subspace identification literature have become standard, despite known statistical efficiency issues. In this paper, we instead reconsider likelihood maximization and develop an optimization based strategy for recovering the latent states and transition parameters. Key to the approach is a two-view reformulation of maximum likelihood estimation for linear dynamical systems that enables the use of global optimization algorithms for matrix factorization. We show that the proposed estimation strategy outperforms widely-used identification algorithms such as subspace identification methods, both in terms of accuracy and runtime.
Citation
M. Karami, M. White, D. Schuurmans, C. Szepesvari. "Multi-view Matrix Factorization for Linear Dynamical System Estimation". NIPS Workshop on Machine Learning and Games, (ed: Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna M. Wallach, Rob Fergus, S. V. N. Vishwanathan, Roman Garnett), pp 7092-7101, December 2017.| Keywords: | |
| Category: | In Conference |
| Web Links: | NeurIPS |
BibTeX
@incollection{Karami+al:NIPS17,
author = {Mahdi Karami and Martha White and Dale Schuurmans and Csaba
Szepesvari},
title = {Multi-view Matrix Factorization for Linear Dynamical System
Estimation},
Editor = {Isabelle Guyon, Ulrike von Luxburg, Samy Bengio, Hanna M. Wallach,
Rob Fergus, S. V. N. Vishwanathan, Roman Garnett},
Pages = {7092-7101},
booktitle = {NIPS Workshop on Machine Learning and Games},
year = 2017,
}Last Updated: February 25, 2020Submitted by Sabina P
