Contextual movement models based on normalizing flows

Research output: Journal contributionsJournal articlesResearchpeer-review

Authors

Movement models predict positions of players (or objects in general) over time and are thus key to analyzing spatiotemporal data as it is often used in sports analytics. Existing movement models are either designed from physical principles or are entirely data-driven. However, the former suffers from oversimplifications to achieve feasible and interpretable models, while the latter relies on computationally costly, from a current point of view, nonparametric density estimations and require maintaining multiple estimators, each responsible for different types of movements (e.g., such as different velocities). In this paper, we propose a unified contextual probabilistic movement model based on normalizing flows. Our approach learns the desired densities by directly optimizing the likelihood and maintains only a single contextual model that can be conditioned on auxiliary variables. Training is simultaneously performed on all observed types of movements, resulting in an effective and efficient movement model. We empirically evaluate our approach on spatiotemporal data from professional soccer. Our findings show that our approach outperforms the state of the art while being orders of magnitude more efficient with respect to computation time and memory requirements.
Original languageEnglish
JournalAStA Advances in Statistical Analysis
Volume107
Issue number1-2
Pages (from-to)51-72
Number of pages22
ISSN1863-8171
DOIs
Publication statusPublished - 03.2023

Bibliographical note

Publisher Copyright:
© 2021, The Author(s).

    Research areas

  • Density estimation, Movement models, Normalizing flows, Soccer data, Spatiotemporal data, Sports analytics
  • Business informatics

Recently viewed

Publications

  1. Global Finite-Time Stabilization of Planar Linear Systems With Actuator Saturation
  2. Analyzing User Journey Data In Digital Health: Predicting Dropout From A Digital CBT-I Intervention
  3. Web-scale extension of RDF knowledge bases from templated websites
  4. Clause identification using entropy guided transformation learning
  5. Experimentally established correlation of friction surfacing process temperature and deposit geometry
  6. Interpreting Strings, Weaving Threads
  7. Generating Energy Optimal Powertrain Force Trajectories with Dynamic Constraints
  8. Analyzing math teacher students' sensitivity for aspects of the complexity of problem oriented mathematics instruction
  9. FaST: A linear time stack trace alignment heuristic for crash report deduplication
  10. What does it mean to be sensitive for the complexity of (problem oriented) teaching?
  11. Improving students’ science text comprehension through metacognitive self-regulation when applying learning strategies
  12. A new way of assessing the interaction of a metallic phase precursor with a modified oxide support substrate as a source of information for predicting metal dispersion
  13. Computing regression statistics from grouped data
  14. Performance analysis for loss systems with many subscribers and concurrent services
  15. Stimulating Computing
  16. TARGET SETTING FOR OPERATIONAL PERFORMANCE IMPROVEMENTS - STUDY CASE -
  17. Integration of laser scanning and projection speckle pattern for advanced pipeline monitoring
  18. Comments on "Tracking Control of Robotic Manipulators With Uncertain Kinematics and Dynamics"
  19. Analysis of long-term statistical data of cobalt flows in the EU
  20. Simulation based optimization of lot sizes for opposing logistic objectives
  21. Gaussian processes for dispatching rule selection in production scheduling
  22. Exploring the limits of graph invariant- and spectrum-based discrimination of (sub)structures.
  23. Learning Analytics with Matlab Grader in Undergraduate Engineering Courses
  24. Neural network-based estimation and compensation of friction for enhanced deep drawing process control
  25. Teaching methods for modelling problems and students’ task-specific enjoyment, value, interest and self-efficacy expectations
  26. Self-regulation in error management training: emotion control and metacognition as mediators of performance effects
  27. Does thinking-aloud affect learning, visual information processing and cognitive load when learning with seductive details as expected from self-regulation perspective?
  28. For a return to the forgotten formula: 'Data 1 + Data 2 > Data 1'
  29. Using Language Learning Resources on YouTube
  30. Cognitive Predictors of Child Second Language Comprehension and Syntactic Learning
  31. A Theoretical Dynamical Noninteracting Model for General Manipulation Systems Using Axiomatic Geometric Structures
  32. Teachers’ use of data from digital learning platforms for instructional design
  33. Dynamic environment modelling and prediction for autonomous systems
  34. Machine Learning and Knowledge Discovery in Databases