LEARNING AND PREDICTION BASED ON A RELATIONAL HIDDEN MARKOV MODEL

Carsten Elfers, Thomas Wagner

2010

Abstract

In this paper we show a novel method on how the well-established hidden markov model and the relational markov model can be combined to the relational hidden markov model to solve currently unrecognized challenging problems of the original models. Our presented methods allows for prediction on different granularity level depending on the validity of the underlying observations. We demonstrate the use of this new method based on a spatio-temporal qualitative representation and validate the approach in the RoboCupSoccer multiagent environment.

Paper Citation

in Harvard Style

Elfers C. and Wagner T. (2010). LEARNING AND PREDICTION BASED ON A RELATIONAL HIDDEN MARKOV MODEL . In Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART, ISBN 978-989-674-021-4, pages 211-216. DOI: 10.5220/0002703202110216

in Bibtex Style

@conference{icaart10,
author={Carsten Elfers and Thomas Wagner},
title={LEARNING AND PREDICTION BASED ON A RELATIONAL HIDDEN MARKOV MODEL},
booktitle={Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,},
year={2010},
pages={211-216},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002703202110216},
isbn={978-989-674-021-4},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 2nd International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,
TI - LEARNING AND PREDICTION BASED ON A RELATIONAL HIDDEN MARKOV MODEL
SN - 978-989-674-021-4
AU - Elfers C.
AU - Wagner T.
PY - 2010
SP - 211
EP - 216
DO - 10.5220/0002703202110216