OmniST: Topology-Invariant Spatiotemporal Forecasting with Guaranteed-Coverage Uncertainty at Unseen Sensor Locations
cs.LG, cs.AI, stat.ML
Submitted: 2026-11-03
Updated: 2026-11-03
Comments: Code and experiments: https://github.com/Praveg432/OmniST . 9 tables, 4 figures. Baselines (E-STGCN, IGNNK) are faithful reimplementations, not the authors' original code
Code: https://github.com/Praveg432/OmniST
License: http://creativecommons.org/licenses/by/4.0/
The gist: Standard spatiotemporal graph convolutional networks need a fixed-size adjacency matrix, coupling model weights to a specific sensor count N…
Terminology
Abstract
Standard spatiotemporal graph convolutional networks need a fixed-size adjacency matrix, coupling model weights to a specific sensor count N; they cannot produce any output at a sensor absent from training. We present OmniST, a topology-invariant forecaster whose computation has zero structural dependency on the number of nodes, and which, composed with inductive split-conformal calibration, gives distribution-free, coverage-guaranteed prediction intervals at arbitrary (including never-seen) sensor locations, preserving that guarantee as sensors are added or removed. Topology invariance follows from coordinate-based tokenization, k-nearest-neighbor attention from geographic coordinates at each forward pass, and per-node temporal processing via a bidirectional GRU. On real Delhi air-quality data (37 stations, 2019-2023), OmniST attains a 30-day PM2.5 MAE of 31.0 ug/m3 and degrades at most 1.1 percent when 40 percent of stations are removed. Its inductive forecasts at unseen stations beat nearest-neighbor interpolation (about 12 percent) and a learned kriging GNN (IGNNK, about 21 percent), and, unlike either, carry a finite-sample coverage guarantee. The guarantee holds at 90 percent or above under 30-50 percent node dropout across four real datasets over three domains (Delhi and Beijing air quality, NOAA GHCN weather, Caltrans PeMSD7-M traffic), and still holds at 325 real sensors (full PeMS-BAY). We also fix a temporal-drift failure: static conformal under-covers over time (about 0.50 late), but a rolling adaptive-conformal estimator restores target coverage (0.89 overall, 0.96 late) at the cost of wider intervals.
Sources
- Conformal Inductive Graph Neural Networks
- Damba-ST: Domain-Adaptive Mamba for Efficient Urban Spatio-Temporal Prediction
- Valid model-free spatial prediction
- More Sensors Only One Field: Rethinking Continual Spatio-Temporal Forecasting
- Relational Conformal Prediction for Correlated Time Series
- Conformal Inference for Time Series over Graphs
- In-Context Residual Calibration for Uncertainty Quantification of Energy Time Series over Graphs
- Valid Conformal Prediction for Dynamic GNNs
- Non-exchangeable Conformal Prediction for Temporal Graph Neural Networks
- Topology-Aware Conformal Prediction for Stream Networks
- Conformalized Gaussian processes for online uncertainty quantification over graphs
Related papers
- A Physics-Regulated Neural Framework for Learning 3D Grain Growth Dynamics
- FlexMoE: Learning One-for-All Multi-Budget Expert Pruning for Flexible MoE Deployment
- OptiMind: Improving Small Language Models for Optimization Modeling through Formulation Guidance and Multi-Turn Learning
- Staying in the Sweet Spot: Boost Reasoning Evolution via Responsive Rollout Difficulty Manipulation
- Examining the Impact of Data and Training Methods on Emergent and Subliminal Misalignment
- Rethinking Black-Box On-Policy Distillation: A Rubric-based Analysis