A prediction interval method for uncertainty quantification of regression models

Conference · Sat May 01 04:00:00 EDT 2021

OSTI ID:1785172

Zhang, Pei ^[1]; Liu, Siyan ^[1]; Lu, Dan ^[1]; Zhang, Guannan ^[1]; Sankaran, Ramanan ^[1]

---

1. ORNL

This paper considers calculation of prediction intervals (PIs) by neural networks (NNs) for quantifying uncertainty in regression tasks, so as to provide fast, accurate and robust emulators to accelerate scientific simulations. We propose a novel method to learn lower and upper bounds of the PI using independent NNs without defining an exclusive loss. Our method requires no distributional assumption, does not introduce extra hyper-parameters, and can effectively identify out-of-distribution samples and quantify their uncertainty. We demonstrate advantages of our method using a benchmark problem and two real-world scientific applications.

• Research Organization: Oak Ridge National Laboratory (ORNL), Oak Ridge, TN (United States)
• Sponsoring Organization: USDOE
• DOE Contract Number: AC05-00OR22725
• OSTI ID: 1785172
• Country of Publication: United States
• Language: English