High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach
Abstract As a multivariate time series, the prediction of curling trajectories is crucial for athletes to devise game strategies. However, the wide prediction range and complex data correlations present significant challenges to this task. This paper puts forward an innovative deep learning approach...
Saved in:
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Nature Portfolio
2025-01-01
|
Series: | Scientific Reports |
Subjects: | |
Online Access: | https://doi.org/10.1038/s41598-025-87933-5 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
_version_ | 1832571673279201280 |
---|---|
author | Yanan Guo Jing Jin Hongyang Zhao Yu Jiang Dandan Li Yi Shen |
author_facet | Yanan Guo Jing Jin Hongyang Zhao Yu Jiang Dandan Li Yi Shen |
author_sort | Yanan Guo |
collection | DOAJ |
description | Abstract As a multivariate time series, the prediction of curling trajectories is crucial for athletes to devise game strategies. However, the wide prediction range and complex data correlations present significant challenges to this task. This paper puts forward an innovative deep learning approach, CasLSTM, by introducing integrated inter-layer memory, and establishes an encoder-predictor curling trajectory forecasting model accordingly. Additionally, tailored training techniques involving non-teacher-forcing, ExMSE loss and incremental multi-trajectory learning are devised to enhance model performance. Notably, the model demonstrates astounding accuracy, achieving sub-1cm average errors over 30m trajectories, outperforming vanilla LSTM by 41.8%. It also showcases robustness across various curling settings, with strict validation metrics on a static test set further verifying precision. Field test results reveal promising predictive capabilities for real-world scenarios as well, exhibiting applicability. The proposed technique liberates data-driven curling stone trajectory prediction from sole reliance on analytical models and tackles key challenges of long sequence forecasting. The presented technologies and insights could also generalize to prediction tasks in other remote trajectories and multivariate time series domains. |
format | Article |
id | doaj-art-a59ed2424f6a4ed5a713586d9812033c |
institution | Kabale University |
issn | 2045-2322 |
language | English |
publishDate | 2025-01-01 |
publisher | Nature Portfolio |
record_format | Article |
series | Scientific Reports |
spelling | doaj-art-a59ed2424f6a4ed5a713586d9812033c2025-02-02T12:24:13ZengNature PortfolioScientific Reports2045-23222025-01-0115111810.1038/s41598-025-87933-5High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approachYanan Guo0Jing Jin1Hongyang Zhao2Yu Jiang3Dandan Li4Yi Shen5Department of Control Science and Engineering, Harbin Institute of TechnologyDepartment of Control Science and Engineering, Harbin Institute of TechnologyCollege of Mechanical and Electrical Engineering, Northeast Forestry UniversityDepartment of Control Science and Engineering, Harbin Institute of TechnologyDepartment of Control Science and Engineering, Harbin Institute of TechnologyDepartment of Control Science and Engineering, Harbin Institute of TechnologyAbstract As a multivariate time series, the prediction of curling trajectories is crucial for athletes to devise game strategies. However, the wide prediction range and complex data correlations present significant challenges to this task. This paper puts forward an innovative deep learning approach, CasLSTM, by introducing integrated inter-layer memory, and establishes an encoder-predictor curling trajectory forecasting model accordingly. Additionally, tailored training techniques involving non-teacher-forcing, ExMSE loss and incremental multi-trajectory learning are devised to enhance model performance. Notably, the model demonstrates astounding accuracy, achieving sub-1cm average errors over 30m trajectories, outperforming vanilla LSTM by 41.8%. It also showcases robustness across various curling settings, with strict validation metrics on a static test set further verifying precision. Field test results reveal promising predictive capabilities for real-world scenarios as well, exhibiting applicability. The proposed technique liberates data-driven curling stone trajectory prediction from sole reliance on analytical models and tackles key challenges of long sequence forecasting. The presented technologies and insights could also generalize to prediction tasks in other remote trajectories and multivariate time series domains.https://doi.org/10.1038/s41598-025-87933-5CurlingDeep learningCascade LSTMInter-layer memoryMultivariate time seriesMulti-step ahead prediction |
spellingShingle | Yanan Guo Jing Jin Hongyang Zhao Yu Jiang Dandan Li Yi Shen High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach Scientific Reports Curling Deep learning Cascade LSTM Inter-layer memory Multivariate time series Multi-step ahead prediction |
title | High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach |
title_full | High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach |
title_fullStr | High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach |
title_full_unstemmed | High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach |
title_short | High-Precision prediction of curling trajectory multivariate time series using the novel CasLSTM approach |
title_sort | high precision prediction of curling trajectory multivariate time series using the novel caslstm approach |
topic | Curling Deep learning Cascade LSTM Inter-layer memory Multivariate time series Multi-step ahead prediction |
url | https://doi.org/10.1038/s41598-025-87933-5 |
work_keys_str_mv | AT yananguo highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach AT jingjin highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach AT hongyangzhao highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach AT yujiang highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach AT dandanli highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach AT yishen highprecisionpredictionofcurlingtrajectorymultivariatetimeseriesusingthenovelcaslstmapproach |