PuSH - Publikationsserver des Helmholtz Zentrums München

Dorigatti, E. ; Schubert, B. ; Bischl, B.* ; Ruegamer, D.*

Frequentist uncertainty quantification in semi-structured neural networks.

In: (Proceedings of The 26th International Conference on Artificial Intelligence and Statistics). 1269 Law St, San Diego, Ca, United States: Jmlr-journal Machine Learning Research, 2023. 18 ( ; 206)
Semi-structured regression (SSR) models jointly learn the effect of structured (tabular) and unstructured (non-tabular) data through additive predictors and deep neural networks (DNNs), respectively. Inference in SSR models aims at deriving confidence intervals for the structured predictor, although current approaches ignore the variance of the DNN estimation of the unstructured effects. This results in an underestimation of the variance of the structured coefficients and, thus, an increase of Type-I error rates. To address this shortcoming, we present here a theoretical framework for structured inference in SSR models that incorporates the variance of the DNN estimate into confidence intervals for the structured predictor. By treating this estimate as a random offset with known variance, our formulation is agnostic to the specific deep uncertainty quantification method employed. Through numerical experiments and a practical application on a medical dataset, we show that our approach results in increased coverage of the true structured coefficients and thus a reduction in Type-I error rate compared to ignoring the variance of the neural network, naive ensembling of SSR models, and a variational inference baseline.
Weitere Metriken?
[➜Einloggen]
Zusatzinfos bearbeiten [➜Einloggen]
Publikationstyp Artikel: Konferenzbeitrag
Korrespondenzautor
ISSN (print) / ISBN 2640-3498
Konferenztitel Proceedings of The 26th International Conference on Artificial Intelligence and Statistics
Quellenangaben Band: 206, Heft: , Seiten: 18 Artikelnummer: , Supplement: ,
Verlag Jmlr-journal Machine Learning Research
Verlagsort 1269 Law St, San Diego, Ca, United States
Nichtpatentliteratur Publikationen
Institut(e) Institute of Computational Biology (ICB)