TimeMIL: Advancing Multivariate Time Series Classification via a Time-aware Multiple Instance Learning

Xiwen Chen, Peijie Qiu, Wenhui Zhu, Huayu Li, Hao Wang, Aristeidis Sotiras, Yalin Wang, Abolfazl Razi
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:7190-7206, 2024.

Abstract

Deep neural networks, including transformers and convolutional neural networks (CNNs), have significantly improved multivariate time series classification (MTSC). However, these methods often rely on supervised learning, which does not fully account for the sparsity and locality of patterns in time series data (e.g., quantification of diseases-related anomalous points in ECG and abnormal detection in signal). To address this challenge, we formally discuss and reformulate MTSC as a weakly supervised problem, introducing a novel multiple-instance learning (MIL) framework for better localization of patterns of interest and modeling time dependencies within time series. Our novel approach, TimeMIL, formulates the temporal correlation and ordering within a time-aware MIL pooling, leveraging a tokenized transformer with a specialized learnable wavelet positional token. The proposed method surpassed 26 recent state-of-the-art MTSC methods, underscoring the effectiveness of the weakly supervised TimeMIL in MTSC. The code is available https://github.com/xiwenc1/TimeMIL.

Cite this Paper


BibTeX
@InProceedings{pmlr-v235-chen24af, title = {{T}ime{MIL}: Advancing Multivariate Time Series Classification via a Time-aware Multiple Instance Learning}, author = {Chen, Xiwen and Qiu, Peijie and Zhu, Wenhui and Li, Huayu and Wang, Hao and Sotiras, Aristeidis and Wang, Yalin and Razi, Abolfazl}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {7190--7206}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/chen24af/chen24af.pdf}, url = {https://proceedings.mlr.press/v235/chen24af.html}, abstract = {Deep neural networks, including transformers and convolutional neural networks (CNNs), have significantly improved multivariate time series classification (MTSC). However, these methods often rely on supervised learning, which does not fully account for the sparsity and locality of patterns in time series data (e.g., quantification of diseases-related anomalous points in ECG and abnormal detection in signal). To address this challenge, we formally discuss and reformulate MTSC as a weakly supervised problem, introducing a novel multiple-instance learning (MIL) framework for better localization of patterns of interest and modeling time dependencies within time series. Our novel approach, TimeMIL, formulates the temporal correlation and ordering within a time-aware MIL pooling, leveraging a tokenized transformer with a specialized learnable wavelet positional token. The proposed method surpassed 26 recent state-of-the-art MTSC methods, underscoring the effectiveness of the weakly supervised TimeMIL in MTSC. The code is available https://github.com/xiwenc1/TimeMIL.} }
Endnote
%0 Conference Paper %T TimeMIL: Advancing Multivariate Time Series Classification via a Time-aware Multiple Instance Learning %A Xiwen Chen %A Peijie Qiu %A Wenhui Zhu %A Huayu Li %A Hao Wang %A Aristeidis Sotiras %A Yalin Wang %A Abolfazl Razi %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-chen24af %I PMLR %P 7190--7206 %U https://proceedings.mlr.press/v235/chen24af.html %V 235 %X Deep neural networks, including transformers and convolutional neural networks (CNNs), have significantly improved multivariate time series classification (MTSC). However, these methods often rely on supervised learning, which does not fully account for the sparsity and locality of patterns in time series data (e.g., quantification of diseases-related anomalous points in ECG and abnormal detection in signal). To address this challenge, we formally discuss and reformulate MTSC as a weakly supervised problem, introducing a novel multiple-instance learning (MIL) framework for better localization of patterns of interest and modeling time dependencies within time series. Our novel approach, TimeMIL, formulates the temporal correlation and ordering within a time-aware MIL pooling, leveraging a tokenized transformer with a specialized learnable wavelet positional token. The proposed method surpassed 26 recent state-of-the-art MTSC methods, underscoring the effectiveness of the weakly supervised TimeMIL in MTSC. The code is available https://github.com/xiwenc1/TimeMIL.
APA
Chen, X., Qiu, P., Zhu, W., Li, H., Wang, H., Sotiras, A., Wang, Y. & Razi, A.. (2024). TimeMIL: Advancing Multivariate Time Series Classification via a Time-aware Multiple Instance Learning. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:7190-7206 Available from https://proceedings.mlr.press/v235/chen24af.html.

Related Material