Goal-Driven Adaptive Sampling Strategies for Machine Learning Models Predicting Fields
This work addresses the problem of computational efficiency in field predictions for researchers and engineers using machine learning with expensive simulations, representing an incremental advancement by extending active learning to more general scenarios.
The paper tackles the challenge of achieving high accuracy in field predictions from expensive simulations with minimal computational cost by proposing an active learning strategy that is agnostic to model architecture, combining Gaussian processes to reduce epistemic error and scalar-field differences, and demonstrates results with high accuracy at significantly lower cost compared to non-active learning approaches for a NASA model uncertainty propagation task.
Machine learning models are widely regarded as a way forward to tackle multi-query challenges that arise once expensive black-box simulations such as computational fluid dynamics are investigated. However, ensuring the desired level of accuracy for a certain task at minimal computational cost, e.g. as few black-box samples as possible, remains a challenges. Active learning strategies are used for scalar quantities to overcome this challenges and different so-called infill criteria exists and are commonly employed in several scenarios. Even though needed in various field an extension of active learning strategies towards field predictions is still lacking or limited to very specific scenarios and/or model types. In this paper we propose an active learning strategy for machine learning models that are capable if predicting field which is agnostic to the model architecture itself. For doing so, we combine a well-established Gaussian process model for a scalar reference value and simultaneously aim at reducing the epistemic model error and the difference between scalar and field predictions. Different specific forms of the above-mentioned approach are introduced and compared to each other as well as only scalar-valued based infill. Results are presented for the NASA common research model for an uncertainty propagation task showcasing high level of accuracy at significantly smaller cost compared to an approach without active learning.