Modeling Conditional Dependencies in Multiagent Trajectories

Research output: Journal contributionsConference article in journalResearchpeer-review

Standard

Modeling Conditional Dependencies in Multiagent Trajectories. / Rudolph, Yannick; Brefeld, Ulf.
In: Proceedings of Machine Learning Research, Vol. 151, 2022, p. 10518-10533.

Research output: Journal contributionsConference article in journalResearchpeer-review

Harvard

APA

Vancouver

Bibtex

@article{f7a4f00425e0440d9fc2d7ab5bcf0b38,
title = "Modeling Conditional Dependencies in Multiagent Trajectories",
abstract = "We study modeling joint densities over sets of random variables (next-step movements of multiple agents) which are conditioned on aligned observations (past trajectories). For this setting, we propose an autoregressive approach to model intra-timestep dependencies, where distributions over joint movements are represented by autoregressive factorizations. In our approach, factors are randomly ordered and estimated with a graph neural network to account for permutation equivariance, while a recurrent neural network encodes past trajectories. We further propose a conditional two-stream attention mechanism, to allow for efficient training of random factorizations. We experiment on trajectory data from professional soccer matches and find that we model low frequency trajectories better than variational approaches.",
keywords = "Informatics, Business informatics",
author = "Yannick Rudolph and Ulf Brefeld",
note = "Funding Information: We would like to thank Marius Lehne at SAP SE for his essential support and help while writing this paper as well as all reviewers for their valuable feedback. Publisher Copyright: Copyright {\textcopyright} 2022 by the author(s); The 25th International Conference on Artificial Intelligence and Statistics - AISTATS 2022, AISTATS 2022 ; Conference date: 28-03-2022 Through 30-03-2022",
year = "2022",
language = "English",
volume = "151",
pages = "10518--10533",
journal = "Proceedings of Machine Learning Research",
issn = "2640-3498",
publisher = "MLResearch Press",
url = "http://aistats.org/aistats2022/",

}

RIS

TY - JOUR

T1 - Modeling Conditional Dependencies in Multiagent Trajectories

AU - Rudolph, Yannick

AU - Brefeld, Ulf

N1 - Conference code: 25

PY - 2022

Y1 - 2022

N2 - We study modeling joint densities over sets of random variables (next-step movements of multiple agents) which are conditioned on aligned observations (past trajectories). For this setting, we propose an autoregressive approach to model intra-timestep dependencies, where distributions over joint movements are represented by autoregressive factorizations. In our approach, factors are randomly ordered and estimated with a graph neural network to account for permutation equivariance, while a recurrent neural network encodes past trajectories. We further propose a conditional two-stream attention mechanism, to allow for efficient training of random factorizations. We experiment on trajectory data from professional soccer matches and find that we model low frequency trajectories better than variational approaches.

AB - We study modeling joint densities over sets of random variables (next-step movements of multiple agents) which are conditioned on aligned observations (past trajectories). For this setting, we propose an autoregressive approach to model intra-timestep dependencies, where distributions over joint movements are represented by autoregressive factorizations. In our approach, factors are randomly ordered and estimated with a graph neural network to account for permutation equivariance, while a recurrent neural network encodes past trajectories. We further propose a conditional two-stream attention mechanism, to allow for efficient training of random factorizations. We experiment on trajectory data from professional soccer matches and find that we model low frequency trajectories better than variational approaches.

KW - Informatics

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=85163100847&partnerID=8YFLogxK

M3 - Conference article in journal

VL - 151

SP - 10518

EP - 10533

JO - Proceedings of Machine Learning Research

JF - Proceedings of Machine Learning Research

SN - 2640-3498

T2 - The 25th International Conference on Artificial Intelligence and Statistics - AISTATS 2022

Y2 - 28 March 2022 through 30 March 2022

ER -

Recently viewed

Activities

  1. Transdisciplinary Evaluation of Alternative Adaptation Strategies Value-Tree Method as a Tool to Integrate Multiple Values of Science, Practice and the General Public into Decision-Making
  2. Material Migrations I Online Lecture Series
  3. It's how, not what we use that matters - Communications Modes in the Internet
  4. Uncertainty and Subjectivity in Provenance Linked Open Data
  5. Using the Multiple Streams Framework and the Multi-Level Perspective to Explain Policy Transformation: The Case of the German Energiewende
  6. Seed dispersal and predation: Interactions, ecosystem functions and services
  7. Continuous Innovation through Modular Upgradeability: How Software Upgrades Affect Consumer Product Valuations
  8. Swarming. Science Fact and Science Fiction of Distributed Intelligence
  9. Validation of school inspection frameworks and methods
  10. Spas in the New Länder: A Transformation with an Uncertain Outcome
  11. Is there a threshold effect of time headway on subjective variables for different velocities?
  12. The Impact of Explicit References in Computer Supported Collaborative Learning: Evidence from Eye Movement Analyses
  13. Alterations of a visual and how they work for and at the boundaries of an interorganizational team: A multimodal exploration
  14. Memory Acts: Memory without Representation.: Theoretical and Methodological Suggestions
  15. Self-tuning of a kalman filter applied in a DC drive and in a kalman-based sensor
  16. Language Policy and Language Learning: New Paradigms and New Challenges - LPLL 2009
  17. Dealing with temporal tensions in drug development processes
  18. Draw the line? How boundary creation behaviors at the end of work relate to recovery and next-day work performance
  19. Es geht um

Publications

  1. Empowering materials processing and performance from data and AI
  2. Intellectual property issues in the use and distribution of remote sensing data
  3. Explaining and controlling for the psychometric properties of computer-generated figural matrix items
  4. Metaphors and Paradigms of the Language Animal—or—The Advantage of seeing “Time Is a Resource” as a Paradigm
  5. Text Comprehension as a Mediator in Solving Mathematical Reality-Based Tasks
  6. Use of Machine-Learning Algorithms Based on Text, Audio and Video Data in the Prediction of Anxiety and Post-Traumatic Stress in General and Clinical Populations
  7. What would Colin say?
  8. Teachers’ use of data from digital learning platforms for instructional design
  9. How Much Home Office is Ideal? A Multi-Perspective Algorithm
  10. Dimension estimates for certain sets of infinite complex continued fractions
  11. Validation of an open source, remote web-based eye-tracking method (WebGazer) for research in early childhood
  12. Is too much help an obstacle? Effects of interactivity and cognitive style on learning with dynamic versus non-dynamic visualizations with narrative explanations
  13. Visualizing the Hidden Activity of Artificial Neural Networks
  14. Q-Adaptive Control of the nonlinear dynamics of the cantilever-sample system of an Atomic Force Microscope
  15. How, when and why do negotiators use reference points?
  16. An observer for sensorless variable valve control in camless internal combustion engines
  17. Using conditional inference trees and random forests to predict the bioaccumulation potential of organic chemicals
  18. Visualization of the Plasma Frequency by means of a Particle Simulation using a Normalized Periodic Model
  19. Assembly Modes of General Planar 3-RPR Parallel Mechanisms when Using the Linear Actuators’ Orientations
  20. Hypertext
  21. Special Issue in Acquisitional Pragmatics in Foreign Language Learning
  22. Distributed robust Gaussian Process regression
  23. Assessment of cognitive load in multimedia learning using dual-task methodology
  24. Diffusion patterns in small vs. large capital markets-the case of value-based management
  25. A MODEL FOR QUANTIFICATION OF SOFTWARE COMPLEXITY
  26. Mathematical relation between extended connectivity and eigenvector coefficients.
  27. Artificial Intelligence in Foreign Language Learning and Teaching
  28. Assembly Theory for Restoring Ecosystem Structure and Functioning