Interpretable and Physics-Supported Machine Learning Model for Sound Transmission Loss Analysis - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2022

Interpretable and Physics-Supported Machine Learning Model for Sound Transmission Loss Analysis

Abstract

Lately, there has been a growing interest in applying Machine Learning and Digital Twins for the speed-up of acoustic simulations. However, the lack of interpretability and physics foundation inhibit the widespread usage of these black-box models by the scientific community. In this article, global sensitivity analysis and feature engineering techniques are leveraged to improve the interpretability and physical consistency of MLbased simulations of the Sound Transmission Loss problem for a variety of plate materials. Computationally efficient sensitivity analysis is obtained via the Mean Decrease in Impurity, which is the byproduct of the training of the Random Forest surrogate models. The resulting sensitivity indices were shown to be similar to the traditional Sobol indices and more accurate than Fourier amplitude sensitivity testing for small datasets. Moreover, introducing basic expert knowledge into the ML inputs helped reduce the surrogate prediction error and interpret the physical meaning of sensitivity indices throughout the frequency spectrum.
Fichier principal
Vignette du fichier
Cunha_ISMA_2022_Full_Paper (1).pdf (806.41 Ko) Télécharger le fichier
Origin Files produced by the author(s)

Dates and versions

hal-03830914 , version 1 (26-10-2022)

Licence

Identifiers

  • HAL Id : hal-03830914 , version 1

Cite

Barbara Zaparoli Cunha, Mohamed Ichchou, Christophe Droz, Abdelmalek Zine, Stéphane Foulard. Interpretable and Physics-Supported Machine Learning Model for Sound Transmission Loss Analysis. ISMA 2022 - International Conference on Noise and Vibration Engineering, Sep 2022, Leuven, Belgium. pp.1-15. ⟨hal-03830914⟩
126 View
63 Download

Share

Gmail Mastodon Facebook X LinkedIn More