ECML PKDD 2025 | 时间序列（Time Series）论文总结

时空探索之旅

发布于 2025-08-07 13:46:11

2890

ECMLP KDD是CCF B类会议。ECML PKDD2025将在2025年9月15号到19号在加葡萄牙波尔图（Porto, Portugal）举行，本文总结了ECML PKDD2025有关时间序列（Time Series）相关文章，共计14篇，其中1-11为Research Track，12-14为ADS Track。

时间序列Topic：预测,分类,异常检测,生成,可解释性等。如有疏漏，欢迎补充！

Research Track1. An Empirical Evaluation of Foundation Models for Multivariate Time Series Classification2. Bridging Neural Networks and Dynamic Time Warping for Adaptive Time Series Classification3. Cross-Domain Conditional Diffusion Models for Time Series Imputation4. Federated Time Series Generation on Feature and Temporally Misaligned Data5. G-GLformer: Transformer with GRU Embedding and Global-Local Attention for Multivariate Time Series Forecasting6. MASCOTS: Model-Agnostic Symbolic COunterfactual explanations for Time Series7. MotiPlus and MotiSet: Discovering the Best Set of Motiflets in Time Series8. Multivariate Time Series Anomaly Prediction Based on Forecasting and Reconstruction Using Transformer with Temporal and Feature-wise Attention9. RandomAD: A Random Kernel-based Anomaly Detector for Time Series10. Right on Time: Revising Time Series Models by Constraining their Explanations11. TSHAP: Fast and Exact SHAP for Explaining Time Series Classification and RegressionADS Track12. Forecasting Irregularly Sampled Time Series with Transformer Encoders13. InterDiff: Synthesizing Financial Time Series with Inter-Stock Correlations via Classifier-Free Guided Diffusion14. Ordinal Aligned Domain Generalization for Sensor-based Time Series Regression

1 An Empirical Evaluation of Foundation Models for Multivariate Time Series Classification

作者：Pinar Sungu Isiacik (University College Dublin)*; Thach Le Nguyen (University College Dublin); Timilehin Aderinola (University College Dublin); Georgiana Ifrim (University College Dublin)

关键词：分类，基础模型

2 Bridging Neural Networks and Dynamic Time Warping for Adaptive Time Series Classification

作者：Jintao Qu (University of Southern California)*; Zichong Wang (Florida International University); Chenhao Wu ( University of Southern California); Wenbin Zhang ( Florida International University); Dongmei Li (Beijing Forestry University)

关键词：分类，自适应

3 Cross-Domain Conditional Diffusion Models for Time Series Imputation

代码：https://github.com/kexin-kxzhang/CD2-TSI

作者：Kexin Zhang (Northwestern University)*; Baoyu Jing (University of Illinois Urbana-Champaign); Selcuk Candan (Arizona State University); Dawei Zhou (Virginia Tech); Qingsong Wen (Squirrel Ai Learning); Han Liu (Northwestern University); Kaize Ding (Northwestern University)

关键词：插补，域适应，扩散模型

4 Federated Time Series Generation on Feature and Temporally Misaligned Data

作者：Zhi Wen Soi (University of Bern); Chenrui Fan (University of Bern); Aditya Shankar (TU Delft); Abel Malan (University of Neuchatel); Lydia Chen (University of Neuchatel)*

关键词：生成，联邦

5 G-GLformer: Transformer with GRU Embedding and Global-Local Attention for Multivariate Time Series Forecasting

作者：Wenjun Yu (Shanghai University of International Business and Economics)*; Jiyanglin Li (Guizhou Key Laboratory of Big Data Statistical Analysis; Guizhou University of Finance and Economics); Wentao Gao (University of South Australia); Niangxi Zhuang (Guangzhou Nanfang College); Wen Li (Shanghai University of International Business and Economics); Shouguo Du (Shanghai Municipal Big Data Center)

关键词：预测，多元时间序列

6 MASCOTS: Model-Agnostic Symbolic COunterfactual explanations for Time Series

作者：Dawid P_udowski (Warsaw University of Technology)*; Francesco Spinnato (University of Pisa); Piotr Wilczy_ski (ETH ZÃ¼rich); Krzysztof Kotowski (KP Labs); Evridiki Ntagiou (European Space Operations Centre); Riccardo Guidotti (University of Pisa); Przemys_aw Biecek (Warsaw University of Technology)

关键词：反事实解释，可解释性

7 MotiPlus and MotiSet: Discovering the Best Set of Motiflets in Time Series

作者：Len Feremans (Universiteit Antwerpen)*; Patrick SchÃ¤fer (Humboldt-University at Berlin); Wannes Meert (KU Leuven)

关键词：模式发现

8 Multivariate Time Series Anomaly Prediction Based on Forecasting and Reconstruction Using Transformer with Temporal and Feature-wise Attention

作者：Chihiro Maru (Chuo University)*; Masato Oguchi (Ochanomizu University); Ichiro Kobayashi (Ochanomizu University)

关键词：异常预测，多变量时间序列预测

9 RandomAD: A Random Kernel-based Anomaly Detector for Time Series

作者：Wenjie Xi (George Mason University)*; Jessica Lin (George Mason University)

关键词：异常检测，卷积

10 Right on Time: Revising Time Series Models by Constraining their Explanations

作者：Maurice Kraus (TU-Darmstadt)*; David Steinmann (TU-Darmstadt); Antonia WÃ¼st (TU-Darmstadt); Andre Kokozinski (TU-Darmstadt); Kristian Kersting (TU-Darmstadt)

关键词：时频交互，“聪明汉斯” 现象

11 TSHAP: Fast and Exact SHAP for Explaining Time Series Classification and Regression

作者：Thach Le Nguyen (University College Dublin)*; Georgiana Ifrim (University College Dublin)

关键词：可解释性，评测

12 Forecasting Irregularly Sampled Time Series with Transformer Encoders

代码：https://github.com/softlab-unimore/ISTF

作者：Riccardo Benassi (University of Modena and Reggio Emilia); Francesco Del Buono (University of Modena and Reggio Emilia); Giacomo Guiduzzi ( University of Modena and Reggio Emilia); Francesco Guerra (University of Modena e Reggio Emilia)*

关键词：不规则时序预测

13 InterDiff: Synthesizing Financial Time Series with Inter-Stock Correlations via Classifier-Free Guided Diffusion

作者：Hou-Wan Long (The Chinese University of Hong Kong)*; Zoufei Tang (Super Quantum Capital Management); Jianhui Zhang (Super Quantum Capital Management); Zhuoyang Zhan (Super Quantum Capital Management); Tao Lu (Southern University of Science and Technology); Xiaoquan Zhang (Tsinghua University)

关键词：股票预测，数据增强，扩散模型

14 Ordinal Aligned Domain Generalization for Sensor-based Time Series Regression

代码：https://github.com/yshi22/OATSDG

作者：Yunchuan Shi (The University of Sydney)*; Wei Li (The University of Sydney); Albert Zomaya (The University of Sydney)

关键词：域泛化，时间序列回归，序数对齐，标签空间移位。