Högskolan i Skövde

his.sePublications
Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Uncertainty of deep learning classifiers: A comparative study across different architectures and quantification methods
University of Skövde, School of Informatics.
2024 (English)Independent thesis Advanced level (degree of Master (One Year)), 10 credits / 15 HE creditsStudent thesis
Abstract [en]

Deep learning models, while powerful, often struggle with uncertainty when faced with noisy or out-of-distribution data, leading to unreliable predictions. Uncertainty in this context refers to the model’s confidence in its predictions, which can be categorized as aleatoric (data-related) or epistemic (model-related). Accurately quantifying this uncertainty is crucial for deploying deep learning systems in critical applications, where over- or under-confident errors can have severe consequences. 

This thesis investigates how different uncertainty quantification methods—specifically Monte Carlo Dropout, Ensemble method, and their combination, Ensemble of Monte Carlo Droput—perform across varying deep learning architectures. Using a synthetic dataset designed to challenge these models, the study examines the effects of network width and depth on uncertainty estimation. Results indicate that wider networks yield sharper, narrower uncertainty boundaries, while deeper networks tend to exhibit uncertainty patterns that better capture the complexity of data. Calibration metrics, including Expected Calibration Error and Brier Score, are used to evaluate the reliability of these models, revealing ongoing challenges in achieving well-calibrated uncertainty estimates. 

The findings offer valuable insights into optimizing both UQ methods and network architectures to enhance the reliability and robustness of deep learning models, contributing to the development of more trustworthy AI systems. 

Place, publisher, year, edition, pages
2024. , p. 56
National Category
Information Systems, Social aspects
Identifiers
URN: urn:nbn:se:his:diva-24581OAI: oai:DiVA.org:his-24581DiVA, id: diva2:1901496
Subject / course
Informationsteknologi
Educational program
Data Science - Master’s Programme
Supervisors
Examiners
Available from: 2024-09-27 Created: 2024-09-27 Last updated: 2024-09-27Bibliographically approved

Open Access in DiVA

fulltext(3980 kB)382 downloads
File information
File name FULLTEXT01.pdfFile size 3980 kBChecksum SHA-512
5c51f35d756cd2c72ccf47851c4928d1bf1138ea980023c43f9a57723d359967c760d82deec496a13d787cc5de5de2d1b800da50d3b38c310b81840b33ee4bc5
Type fulltextMimetype application/pdf

By organisation
School of Informatics
Information Systems, Social aspects

Search outside of DiVA

GoogleGoogle Scholar
Total: 384 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 225 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • apa-cv
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf