Progressive Multi-Fidelity Learning

Last updated on Jan 26, 2026
code

We propose a progressive multi-fidelity surrogate modeling paradigm, designed to incorporate new datasets of varying modalities as they become available. This allows to progressively enhance prediction accuracy and reduce uncertainty, while effectively ensuring knowledge retention across updates.

Contributors: Paolo Conti.

Deep Learning Multi-Fidelity
Paolo Conti
Paolo Conti
Principal Research Scientist

My research interests include scientific machine learning, multi-fidelity and reduced-order modeling.

Publications

Progressive multi-fidelity learning with neural networks for physical system predictions
Highly accurate datasets from numerical or physical experiments are often expensive and time-consuming to acquire, posing a significant …
Paolo Conti, Mengwu Guo, Attilio Frangi, Andrea Manzoni
PDF Cite Code Project DOI
Progressive multi-fidelity learning with neural networks for physical system predictions
Soft sensor design using hierarchical multi-fidelity modeling with bayesian optimization for input variable selection
Soft sensors, or inferential sensors, are crucial in quality and process control systems because they allow for e!cient, online …
Johannes Lips, Hendrik Lens, Paolo Conti
PDF Cite Code Project DOI
Soft sensor design using hierarchical multi-fidelity modeling with bayesian optimization for input variable selection