ConTSG-Bench: A Unified Benchmark for Conditional Time Series Generation

Shaocheng Lan, Shuqi Gu, Zhangzhi Xiong, Kan Ren
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:62454-62493, 2026.

Abstract

Conditional time series generation plays a critical role in addressing data scarcity and enabling causal analysis in real-world applications. Despite its increasing importance, the field lacks a standardized and systematic benchmarking framework for evaluating generative models across diverse conditions. To address this gap, we introduce the Conditional Time Series Generation Benchmark (ConTSG-Bench). ConTSG-Bench comprises a suite of large-scale, well-aligned datasets spanning diverse conditioning modalities and levels of semantic abstraction, enabling systematic evaluation of representative generation methods across these dimensions with a comprehensive suite of metrics for generation fidelity and condition adherence. Both the quantitative benchmarking and in-depth analyses of conditional generation behaviors have revealed the traits and limitations of the current approaches, highlighting critical challenges and promising research directions, particularly with respect to precise structural controllability and downstream task utility under complex conditions.

Cite this Paper


BibTeX
@InProceedings{pmlr-v306-lan26a, title = {{C}on{TSG}-Bench: A Unified Benchmark for Conditional Time Series Generation}, author = {Lan, Shaocheng and Gu, Shuqi and Xiong, Zhangzhi and Ren, Kan}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {62454--62493}, 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/lan26a/lan26a.pdf}, url = {https://proceedings.mlr.press/v306/lan26a.html}, abstract = {Conditional time series generation plays a critical role in addressing data scarcity and enabling causal analysis in real-world applications. Despite its increasing importance, the field lacks a standardized and systematic benchmarking framework for evaluating generative models across diverse conditions. To address this gap, we introduce the Conditional Time Series Generation Benchmark (ConTSG-Bench). ConTSG-Bench comprises a suite of large-scale, well-aligned datasets spanning diverse conditioning modalities and levels of semantic abstraction, enabling systematic evaluation of representative generation methods across these dimensions with a comprehensive suite of metrics for generation fidelity and condition adherence. Both the quantitative benchmarking and in-depth analyses of conditional generation behaviors have revealed the traits and limitations of the current approaches, highlighting critical challenges and promising research directions, particularly with respect to precise structural controllability and downstream task utility under complex conditions.} }
Endnote
%0 Conference Paper %T ConTSG-Bench: A Unified Benchmark for Conditional Time Series Generation %A Shaocheng Lan %A Shuqi Gu %A Zhangzhi Xiong %A Kan Ren %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-lan26a %I PMLR %P 62454--62493 %U https://proceedings.mlr.press/v306/lan26a.html %V 306 %X Conditional time series generation plays a critical role in addressing data scarcity and enabling causal analysis in real-world applications. Despite its increasing importance, the field lacks a standardized and systematic benchmarking framework for evaluating generative models across diverse conditions. To address this gap, we introduce the Conditional Time Series Generation Benchmark (ConTSG-Bench). ConTSG-Bench comprises a suite of large-scale, well-aligned datasets spanning diverse conditioning modalities and levels of semantic abstraction, enabling systematic evaluation of representative generation methods across these dimensions with a comprehensive suite of metrics for generation fidelity and condition adherence. Both the quantitative benchmarking and in-depth analyses of conditional generation behaviors have revealed the traits and limitations of the current approaches, highlighting critical challenges and promising research directions, particularly with respect to precise structural controllability and downstream task utility under complex conditions.
APA
Lan, S., Gu, S., Xiong, Z. & Ren, K.. (2026). ConTSG-Bench: A Unified Benchmark for Conditional Time Series Generation. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:62454-62493 Available from https://proceedings.mlr.press/v306/lan26a.html.

Related Material