Distribution Preserving Multiple Hypotheses Prediction for Uncertainty Modeling

Leemann T, Sackmann M, Thielecke J, Hofmann U (2021)

Publication Language: English

Publication Type: Conference contribution, Conference Contribution

Publication year: 2021

Publisher: ESANN

Pages Range: 523-528

Conference Proceedings Title: Proceedings of the 29th European Symposium on Artificial Neural Networks

Event location: online BE

ISBN: 978287587082-7

URI: https://www.esann.org/sites/default/files/proceedings/2021/ES2021-16.pdf

DOI: 10.14428/esann/2021.es2021-16

Abstract

Many supervised machine learning tasks, such as future state prediction in dynamical systems, require precise modeling of a forecast’s uncertainty. The Multiple Hypotheses Prediction (MHP) approach addresses this problem by providing several hypotheses that represent possible outcomes. Unfortunately, with the common l2 loss function, these hypotheses do not preserve the data distribution’s characteristics. We propose an alternative loss for distribution preserving MHP and review relevant theorems supporting our claims. Furthermore, we empirically show that our approach yields more representative hypotheses on a synthetic and a real-world motion prediction data set. The outputs of the proposed method can directly be used in sampling-based Monte-Carlo methods.

Authors with CRIS profile

Moritz Sackmann Lehrstuhl für Informationstechnik mit dem Schwerpunkt Kommunikationselektronik (Stiftungslehrstuhl) Jörn Thielecke Professur für Informationstechnik (Schwerpunkt Ortsbestimmung und Navigation)

Involved external institutions

Audi AG DE Germany (DE) Eberhard Karls Universität Tübingen DE Germany (DE)

How to cite

APA:

Leemann, T., Sackmann, M., Thielecke, J., & Hofmann, U. (2021). Distribution Preserving Multiple Hypotheses Prediction for Uncertainty Modeling. In Proceedings of the 29th European Symposium on Artificial Neural Networks (pp. 523-528). online, BE: ESANN.

MLA:

Leemann, Tobias, et al. "Distribution Preserving Multiple Hypotheses Prediction for Uncertainty Modeling." Proceedings of the European Symposium on Artificial Neural Networks (ESANN), online ESANN, 2021. 523-528.

BibTeX: Download