Publications

You can also find my articles on my Google Scholar profile.

Journal Articles

Adapting to the Stream: An Instance-Attention GNN Method for Irregular Multivariate Time Series Data

Published in Frontiers of Computer Science, Vol. 19 (Issue 8), 2025

We propose an instance-attention graph neural network (IA-GNN) that learns robust representations for irregular and multivariate time-series streams. By modelling cross-timestamp relations as a dynamic graph and applying fine-grained instance attention, the method achieves state-of-the-art accuracy while remaining interpretable.

Recommended citation: **Kun Han**, Abigail M.Y. Koay, Ryan K.L. Ko, Weitong Chen & Miao Xu (2025). “Adapting to the Stream: An Instance-Attention GNN Method for Irregular Multivariate Time Series Data.” *Frontiers of Computer Science*, 19(8):198340. https://doi.org/10.1007/s11704-024-40449-z
Download Paper | Download Slides

Conference Papers

HITS: Hierarchical Interpretable Time Series Classification via Multiple Instance Learning

Published in IJCNN 2025 – International Joint Conference on Neural Networks, 2025

HITS introduces a two-level MIL framework that combines variable-level and temporal-patch attention, achieving state-of-the-art accuracy on irregular multivariate time-series benchmarks while delivering clear, human-readable explanations.

Recommended citation: Kun Han, Abigail M.Y. Koay, Ryan K.L. Ko, Weitong Chen & Miao Xu (2025). “HITS: Hierarchical Interpretable Time Series Classification via Multiple Instance Learning.” *Proceedings of the International Joint Conference on Neural Networks (IJCNN 2025)*. Rome, Italy.
Download Paper | Download Slides

Mining Irregular Time Series Data with Noisy Labels: A Risk Estimation Approach

Published in Australasian Database Conference (ADC 2025), 2025

We introduce a principled risk-estimation framework for mining irregular time-series data with noisy labels. By modelling per-instance noise and calibrating risk bounds, our method delivers robust classification while providing uncertainty estimates that aid downstream decision-making.

Recommended citation: Kun Han, Abigail M.Y. Koay, Ryan K.L. Ko, Weitong Chen & Miao Xu (2025). “Mining Irregular Time Series Data with Noisy Labels: A Risk Estimation Approach.” *Proceedings of the Australasian Database Conference (ADC 2025)*, pp. 293–307. Springer, Singapore.
Download Paper | Download Slides

Investigating Active Positive-Unlabeled Learning with Deep Networks

Published in AI 2021 – 34ᵗʰ Australasian Joint Conference on Artificial Intelligence, 2022

We propose an active learning framework that combines positive–unlabeled (PU) learning with deep networks. By actively querying the most informative negatives, our method reduces annotation cost while maintaining competitive performance on noisy real-world data.

Recommended citation: Kun Han, Weitong Chen, & Miao Xu. (2022). “Investigating Active Positive-Unlabeled Learning with Deep Networks.” *Proceedings of the 34ᵗʰ Australasian Joint Conference on Artificial Intelligence (AI 2021)*, pp. 607–618. https://doi.org/10.1007/978-3-030-97546-3_49
Download Paper | Download Slides