DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

Estimating Uncertainty of Deep Learning Multi-Label Classifications using Laplace Approximation

Rewicki, Ferdinand (2021) Estimating Uncertainty of Deep Learning Multi-Label Classifications using Laplace Approximation. Master's, Friedrich-Schiller-Universität Jena.

[img] PDF


With the huge successes of deep learning and its application in critical areas such as medical diagnosis or autonomous driving and in fields with noisy and very varying data such as remote sensing, the need for reliable confidence statements about such model's predictions becomes apparent. Therefore, uncertainty estimation methods for neural networks have raised rising interest in the machine learning community. While various methods for regression and multi-class classification tasks have been published, the field of multi-label classification has hardly been considered yet. In this work, we derive the Kronecker-factored Laplace approximation in the multi-label setting, a method to approximate the intractable posterior distribution over the parameters of neural networks. We employ this method in the remote sensing domain and estimate the model uncertainty of eight deep neural networks that have been trained on an aerial scene classification dataset. By comparing the probabilistic classifiers to their deterministic counterparts, we evaluate the potential for using the uncertainty estimates to improve the calibration of those classifiers as well as the out-of-distribution detection. We found that we can improve the calibration for overconfident classifiers whereas for underconfident ones, this method might not be beneficial. Furthermore, the ability to improve the separation from in- and out-of-distribution data seems to be depending on the depth of the neural network within one model family.

Item URL in elib:https://elib.dlr.de/144744/
Document Type:Thesis (Master's)
Title:Estimating Uncertainty of Deep Learning Multi-Label Classifications using Laplace Approximation
AuthorsInstitution or Email of AuthorsAuthor's ORCID iDORCID Put Code
Rewicki, FerdinandUNSPECIFIEDhttps://orcid.org/0000-0003-2264-9495UNSPECIFIED
Date:August 2021
Refereed publication:No
Open Access:Yes
Gold Open Access:No
In ISI Web of Science:No
Number of Pages:87
Keywords:Multi-Label Classification, Bayesian Deep Neural Networks, Uncertainty Estimation, Laplace Approximation, Remote Sensing
Institution:Friedrich-Schiller-Universität Jena
Department:Fakultät für Mathematik und Informatik, Lehrstuhl für Theoretische Informatik II
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Space
HGF - Program Themes:other
DLR - Research area:Raumfahrt
DLR - Program:R - no assignment
DLR - Research theme (Project):R - no assignment
Location: Jena
Institutes and Institutions:Institute of Data Science > Datamangagement and Analysis
Deposited By: Rewicki, Ferdinand
Deposited On:27 Oct 2021 15:31
Last Modified:27 Oct 2021 15:31

Repository Staff Only: item control page

Help & Contact
electronic library is running on EPrints 3.3.12
Website and database design: Copyright © German Aerospace Center (DLR). All rights reserved.