Where to Measure: Epistemic Uncertainty-Based Sensor Placement with ConvCNPs
This addresses sensor placement for environmental and climate modeling, though it appears incremental as it builds on existing ConvCNP methods with a modified uncertainty component.
The paper tackles the problem of suboptimal sensor placement in spatio-temporal systems by proposing a new acquisition function based on expected reduction in epistemic uncertainty, which preliminary results suggest reduces model error more effectively than approaches using total predictive uncertainty.
Accurate sensor placement is critical for modeling spatio-temporal systems such as environmental and climate processes. Neural Processes (NPs), particularly Convolutional Conditional Neural Processes (ConvCNPs), provide scalable probabilistic models with uncertainty estimates, making them well-suited for data-driven sensor placement. However, existing approaches rely on total predictive uncertainty, which conflates epistemic and aleatoric components, that may lead to suboptimal sensor selection in ambiguous regions. To address this, we propose expected reduction in epistemic uncertainty as a new acquisition function for sensor placement. To enable this, we extend ConvCNPs with a Mixture Density Networks (MDNs) output head for epistemic uncertainty estimation. Preliminary results suggest that epistemic uncertainty driven sensor placement more effectively reduces model error than approaches based on overall uncertainty.