This repository contains the Python based implementation of the Multivariate Sensitivity Adaptive (MVSA) Polynomial Chaos Expansion (PCE) method, which is described in the paper "Multivariate sensitivity-adaptive polynomial chaos expansion for high-dimensional surrogate modeling and uncertainty quantification" by D. Loukrezis, E. Diehl, and H. De Gersem, available at https://doi.org/10.1016/j.apm.2024.115746. Please cite this work, in case you use and/or refer to the MVSA PCE method and/or related software.
@article{loukrezis2025multivariate, title={Multivariate sensitivity-adaptive polynomial chaos expansion for high-dimensional surrogate modeling and uncertainty quantification}, author={Loukrezis, Dimitrios and Diehl, Eric and De Gersem, Herbert}, journal={Applied Mathematical Modelling}, volume={137}, pages={115746}, year={2025}, publisher={Elsevier} }
The present software and the related examples rely partially on the OpenTURNS C++/Python library.
- http://www.openturns.org/
- Open TURNS: An industrial software for uncertainty quantification in simulation, https://arxiv.org/abs/1501.05242
Please note that the induction motor example uses version 1.1.0 of the gym-electric-motor package. Unfortunately, later versions result in errors.