Active Learning for Finite Element Simulations with Adaptive Non-Stationary Kernel Function

Abstract: Simulating complex physical processes across a domain of input parameters can be very computationally expensive. Multi-fidelity surrogate modeling can resolve this issue by integrating cheaper simulations with the expensive ones in order to obtain better predictions at a reasonable cost. We are specifically interested in computer experiments that involve the use of finite element methods with a real-valued tuning parameter that determines the fidelity of the numerical output. In these cases, integrating this fidelity parameter in the analysis enables us to make inference on fidelity levels that have not been observed yet. Such models have been developed, and we propose a new adaptive non-stationary kernel function which more accurately reflects the behavior of computer simulation outputs. In addition, we aim to create a sequential design based on the integrated mean squared prediction error (IMSPE) to identify the best design points across input parameters and fidelity parameter, while taking into account the computational cost associated with the fidelity parameter. We illustrate this methodology through synthetic examples and applications to finite element analysis. An $\textsf{R}$ package for the proposed methodology is provided in an open repository.