Conformalized-DeepONet: A Distribution-Free Framework for Uncertainty Quantification in Deep Operator Networks
About
In this paper, we adopt conformal prediction, a distribution-free uncertainty quantification (UQ) framework, to obtain confidence prediction intervals with coverage guarantees for Deep Operator Network (DeepONet) regression. Initially, we enhance the uncertainty quantification frameworks (B-DeepONet and Prob-DeepONet) previously proposed by the authors by using split conformal prediction. By combining conformal prediction with our Prob- and B-DeepONets, we effectively quantify uncertainty by generating rigorous confidence intervals for DeepONet prediction. Additionally, we design a novel Quantile-DeepONet that allows for a more natural use of split conformal prediction. We refer to this distribution-free effective uncertainty quantification framework as split conformal Quantile-DeepONet regression. Finally, we demonstrate the effectiveness of the proposed methods using various ordinary, partial differential equation numerical examples, and multi-fidelity learning.
Related benchmarks
| Task | Dataset | Result | Rank | |
|---|---|---|---|---|
| Air Quality Prediction | air quality | FC0.565 | 9 | |
| Energy forecasting | Energy Demand | FC49.6 | 9 | |
| Uncertainty Quantification | Reg-GP1D Global Heterogeneity Gaussian process simulations (test) | Fidelity Coverage (FC)52.7 | 9 | |
| Uncertainty Quantification | AR-GP1D Spectral Heterogeneity Gaussian process simulations (test) | FC0.206 | 9 | |
| Uncertainty Quantification | AR-GP2D Local Heterogeneity (test) | FC0.00e+0 | 9 | |
| Weather forecasting | Weather-ERA5 | FC Score0.00e+0 | 9 |