Collapsed Amortized Variational Inference for Switching Nonlinear Dynamical Systems
Machine Learning
2020-02-12 v2 Machine Learning
Abstract
We propose an efficient inference method for switching nonlinear dynamical systems. The key idea is to learn an inference network which can be used as a proposal distribution for the continuous latent variables, while performing exact marginalization of the discrete latent variables. This allows us to use the reparameterization trick, and apply end-to-end training with stochastic gradient descent. We show that the proposed method can successfully segment time series data, including videos and 3D human pose, into meaningful ``regimes'' by using the piece-wise nonlinear dynamics.
Cite
@article{arxiv.1910.09588,
title = {Collapsed Amortized Variational Inference for Switching Nonlinear Dynamical Systems},
author = {Zhe Dong and Bryan A. Seybold and Kevin P. Murphy and Hung H. Bui},
journal= {arXiv preprint arXiv:1910.09588},
year = {2020}
}