State Tracking in the Presence of Heavy-tailed Observations

Yaman Kindap

2021

Abstract

In this paper, we define a state-space model with discrete latent states and a multivariate heavy-tailed observation density for applications in tracking the state of a system with observations including extreme deviations from the median. We use a Gaussian distribution with an unknown variance parameter which has a Gamma distribution prior depending on the state of the system to model the observation density. The key contribution of the paper is the theoretical formulation of such a state-space model which makes use of scale mixtures of Gaussians to yield an exact inference method. We derive the framework for estimation of the states and how to estimate the parameters of the model. We demonstrate the performance of the model on synthetically generated data sets.

Paper Citation

in Harvard Style

Kindap Y. (2021). State Tracking in the Presence of Heavy-tailed Observations.In Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM, ISBN 978-989-758-486-2, pages 135-142. DOI: 10.5220/0010150601350142

in Bibtex Style

@conference{icpram21,
author={Yaman Kindap},
title={State Tracking in the Presence of Heavy-tailed Observations},
booktitle={Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,},
year={2021},
pages={135-142},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0010150601350142},
isbn={978-989-758-486-2},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 10th International Conference on Pattern Recognition Applications and Methods - Volume 1: ICPRAM,
TI - State Tracking in the Presence of Heavy-tailed Observations
SN - 978-989-758-486-2
AU - Kindap Y.
PY - 2021
SP - 135
EP - 142
DO - 10.5220/0010150601350142