PATRA: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering

Junkai Lu, Peng Chen, Xingjian Wu, Yang Shu, Chenjuan Guo, Christian S. Jensen, Bin Yang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:80825-80845, 2026.

Abstract

Time series reasoning demands both the perception of complex dynamics and logical depth. However, existing LLM-based approaches exhibit two limitations: they often treat time series merely as text or images, failing to capture the patterns like trends and seasonalities needed to answer specific questions; and when trained on a mix of simple and complex tasks, simpler objectives often dominate the learning process, hindering the development of deep reasoning capabilities. To address these limitations, we propose the Pattern-Aware Alignment and Balanced Reasoning model (PATRA), introducing a pattern-aware mechanism that extracts trend and seasonality patterns from time series to achieve deep alignment. Furthermore, we design a task-aware balanced reward to harmonize learning across tasks of varying difficulty, incentivizing the generation of coherent Chains of Thought. Extensive experiments show that PATRA outperforms strong baselines across diverse Time Series Question Answering (TSQA) tasks, demonstrating superior cross-modal understanding and reasoning capability.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-lu26b, title = {{PATRA}: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering}, author = {Lu, Junkai and Chen, Peng and Wu, Xingjian and Shu, Yang and Guo, Chenjuan and Jensen, Christian S. and Yang, Bin}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {80825--80845}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/lu26b/lu26b.pdf}, url = {https://proceedings.mlr.press/v306/lu26b.html}, abstract = {Time series reasoning demands both the perception of complex dynamics and logical depth. However, existing LLM-based approaches exhibit two limitations: they often treat time series merely as text or images, failing to capture the patterns like trends and seasonalities needed to answer specific questions; and when trained on a mix of simple and complex tasks, simpler objectives often dominate the learning process, hindering the development of deep reasoning capabilities. To address these limitations, we propose the Pattern-Aware Alignment and Balanced Reasoning model (PATRA), introducing a pattern-aware mechanism that extracts trend and seasonality patterns from time series to achieve deep alignment. Furthermore, we design a task-aware balanced reward to harmonize learning across tasks of varying difficulty, incentivizing the generation of coherent Chains of Thought. Extensive experiments show that PATRA outperforms strong baselines across diverse Time Series Question Answering (TSQA) tasks, demonstrating superior cross-modal understanding and reasoning capability.} }
Endnote
%0 Conference Paper %T PATRA: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering %A Junkai Lu %A Peng Chen %A Xingjian Wu %A Yang Shu %A Chenjuan Guo %A Christian S. Jensen %A Bin Yang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-lu26b %I PMLR %P 80825--80845 %U https://proceedings.mlr.press/v306/lu26b.html %V 306 %X Time series reasoning demands both the perception of complex dynamics and logical depth. However, existing LLM-based approaches exhibit two limitations: they often treat time series merely as text or images, failing to capture the patterns like trends and seasonalities needed to answer specific questions; and when trained on a mix of simple and complex tasks, simpler objectives often dominate the learning process, hindering the development of deep reasoning capabilities. To address these limitations, we propose the Pattern-Aware Alignment and Balanced Reasoning model (PATRA), introducing a pattern-aware mechanism that extracts trend and seasonality patterns from time series to achieve deep alignment. Furthermore, we design a task-aware balanced reward to harmonize learning across tasks of varying difficulty, incentivizing the generation of coherent Chains of Thought. Extensive experiments show that PATRA outperforms strong baselines across diverse Time Series Question Answering (TSQA) tasks, demonstrating superior cross-modal understanding and reasoning capability.
APA
Lu, J., Chen, P., Wu, X., Shu, Y., Guo, C., Jensen, C.S. & Yang, B.. (2026). PATRA: Pattern-Aware Alignment and Balanced Reasoning for Time Series Question Answering. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:80825-80845 Available from https://proceedings.mlr.press/v306/lu26b.html.

Related Material