TY - CHAP

T1 - Masked Autoencoder Pretraining for Event Classification in Elite Soccer

AU - Rudolph, Yannick

AU - Brefeld, Ulf

N1 - Conference code: 10

PY - 2024/2/26

Y1 - 2024/2/26

N2 - We show that pretraining transformer models improves the performance on supervised classification of tracking data from elite soccer. Specifically, we propose a novel self-supervised masked autoencoder for multiagent trajectories. In contrast to related work, our approach is significantly simpler, has no necessity for handcrafted features and inherently allows for permutation invariance in downstream tasks.

AB - We show that pretraining transformer models improves the performance on supervised classification of tracking data from elite soccer. Specifically, we propose a novel self-supervised masked autoencoder for multiagent trajectories. In contrast to related work, our approach is significantly simpler, has no necessity for handcrafted features and inherently allows for permutation invariance in downstream tasks.

KW - Factorized transformer architecture

KW - Masked autoencoder

KW - Multiagent trajectories

KW - Self-supervised learning

KW - Soccer

KW - Tracking data

KW - Informatics

KW - Business informatics

UR - http://www.scopus.com/inward/record.url?scp=85187650562&partnerID=8YFLogxK

UR - https://www.mendeley.com/catalogue/df720300-8c0a-37b1-a340-cb810b405c19/

U2 - 10.1007/978-3-031-53833-9_3

DO - 10.1007/978-3-031-53833-9_3

M3 - Article in conference proceedings

AN - SCOPUS:85187650562

SN - 978-3-031-53832-2

T3 - Communications in Computer and Information Science

SP - 24

EP - 35

BT - Machine Learning and Data Mining for Sports Analytics

A2 - Brefeld, Ulf

A2 - Davis, Jesse

A2 - Van Haaren, Jan

A2 - Zimmermann, Albrecht

PB - Springer Nature Switzerland AG

CY - Cham

T2 - 10th International Workshop on Machine Learning and Data Mining for Sports Analytics - MLSA 2023

Y2 - 18 September 2023 through 18 September 2023

ER -