Fast Calibration for Computer Models with Massive Physical Observations

TL;DR

This paper introduces a fast, scalable calibration method for computer models using subsampling, significantly reducing computational complexity while maintaining accuracy, and providing uncertainty quantification.

Contribution

A novel two-step subsampling algorithm for computer model calibration that is computationally efficient and theoretically justified.

Findings

Reduced computational complexity compared to existing methods.

Proven consistency and asymptotic normality of the estimator.

Validated effectiveness through simulations and real-case studies.

Abstract

Computer model calibration is a crucial step in building a reliable computer model. In the face of massive physical observations, a fast estimation for the calibration parameters is urgently needed. To alleviate the computational burden, we design a two-step algorithm to estimate the calibration parameters by employing the subsampling techniques. Compared with the current state-of-the-art calibration methods, the complexity of the proposed algorithm is greatly reduced without sacrificing too much accuracy. We prove the consistency and asymptotic normality of the proposed estimator. The form of the variance of the proposed estimation is also presented, which provides a natural way to quantify the uncertainty of the calibration parameters. The obtained results of two numerical simulations and two real-case studies demonstrate the advantages of the proposed method.