Publications

Recent publications, preprints, and manuscripts from SIGNAL Lab.

2026 publications

Tensor Topic Modeling via HOSVD

Yating Liu and Claire Donnat.

Journal of Machine Learning Research · Accepted, forthcoming · 2026

Canonical Correlation Analysis as Reduced Rank Regression in High Dimensions

Claire Donnat and Elena Tuzhilina.

Journal of Machine Learning Research, 27(160):1–84 · 2026

Efficient Canonical Correlation Analysis with Sparsity

Zixuan Wu, Coralie Rousseau, Elena Tuzhilina, and Claire Donnat.

Electronic Journal of Statistics · 2026

SpeedCP: Fast Kernel-based Conditional Conformal Prediction

Yating Liu, Yeo Jin Jung, Zixuan Wu, So Won Jeong, and Claire Donnat.

International Conference on Machine Learning (ICML) · 2026

Sparse Topic Modeling via Spectral Decomposition and Thresholding

Huy Tran, Yating Liu, and Claire Donnat.

Journal of Machine Learning Research, 27(59):1–76 · 2026

Denoising over networks with applications to partially observed epidemics

Claire Donnat, Olga Klopp, and Nicolas Verzelen.

Computational Statistics & Data Analysis, 215 · 2026

Preprints & manuscripts

Bacterial vitamin sharing emerges from a balance between release and uptake

Freddy Bunbury, Thomas Janas, Paige Mullen, Kaylie Scorza, Sagnik Ghosh, Jeffrey Zhang, Madhav Mani, Catherine A. Pfister, Claire Donnat, and Seppe Kuehn.

bioRxiv · 2026

SAFE-DRIFT: Data Selection for Supervised Fine-Tuning with Controllable Off-Target Drifts

Yeo Jin Jung, Yating Liu, Lalchland Pandia, and Claire Donnat.

Manuscript · Submitted to NeurIPS 2026

Bias, Measurement Error, and Double-Dipping: When Can GNN Convolutions Help Brain Connectome Prediction?

Tommaso Castellani, Jiaqi Li, and Claire Donnat.

Manuscript · Submitted to NeurIPS 2026

Topic Modeling Reveals Thermally Partitioned and Taxonomically Distinct Microbial Subcommunities across Prokaryotes and Phytoplankton in the Laurentian Great Lakes

María D. Hernández Limón, Claire Donnat, Freddy Bunbury, and Maureen L. Coleman.

bioRxiv · 2026

Semi-Supervised Learning on Graphs using Graph Neural Networks

Juntong Chen, Claire Donnat, Olga Klopp, and Johannes Schmidt-Hieber.

arXiv · 2026 · Under review at JASA

Joint learning of a network of linear dynamical systems via total variation penalization

Claire Donnat, Olga Klopp, and Hemant Tyagi.

arXiv · 2025, revised 2026 · Under review at JMLR

Filtering with Confidence: When Data Augmentation Meets Conformal Prediction

Zixuan Wu, So Won Jeong, Yating Liu, Yeo Jin Jung, and Claire Donnat.

arXiv · 2025, revised 2026

LOBSTUR: A Local Bootstrap Framework for Tuning Unsupervised Representations in Graph Neural Networks

So Won Jeong and Claire Donnat.

arXiv · 2025

Graph Topic Modeling for Documents with Spatial or Covariate Dependencies

Yeo Jin Jung and Claire Donnat.

arXiv · 2024, revised 2025

2025 publications

Expediting Field-Effect Transistor Chemical Sensor Design with Neuromorphic Spiking Graph Neural Networks

Rodrigo P. Ferreira, Rui Ding, Fengxue Zhang, Haihui Pu, Claire Donnat, Yuxin Chen, and Junhong Chen.

Molecular Systems Design & Engineering, 10:345–356 · 2025

Understanding the Effect of GCN Convolutions in Regression Tasks

Juntong Chen, Johannes Schmidt-Hieber, Claire Donnat, and Olga Klopp.

AISTATS, Proceedings of Machine Learning Research, 258:4573–4581 · 2025

The Generalized Elastic Net for Least Squares Regression With Network-Aligned Signal and Correlated Design

Huy Tran, Sansen Wei, and Claire Donnat.

IEEE Transactions on Signal and Information Processing over Networks, 11:1459–1472 · 2025