The Misclassification Likelihood Matrix: Some Classes Are More Likely To Be Misclassified Than Others

Simple item page
dc.contributor.authorSikar, Danielen_US
dc.contributor.authorGarcez, Artur d'Avilaen_US
dc.contributor.authorBloomfield, Robinen_US
dc.contributor.authorWeyde, Tillmanen_US
dc.contributor.authorPeeroo, Kaleemen_US
dc.contributor.authorSingh, Namanen_US
dc.contributor.authorHutchinson, Maeveen_US
dc.contributor.authorLaksono, Danyen_US
dc.contributor.authorReljan-Delaney, Mirelaen_US
dc.contributor.editorHunter, Daviden_US
dc.contributor.editorSlingsby, Aidanen_US
dc.date.accessioned2024-09-09T05:45:44Z
dc.date.available2024-09-09T05:45:44Z
dc.date.issued2024
dc.description.abstractThis study introduces the Misclassification Likelihood Matrix (MLM) as a novel tool for quantifying the reliability of neural network predictions under distribution shifts. The MLM is obtained by leveraging softmax outputs and clustering techniques to measure the distances between the predictions of a trained neural network and class centroids. By analyzing these distances, the MLM provides a comprehensive view of the model's misclassification tendencies, enabling decision-makers to identify the most common and critical sources of errors. The MLM allows for the prioritization of model improvements and the establishment of decision thresholds based on acceptable risk levels. The approach is evaluated on the MNIST dataset using a Convolutional Neural Network (CNN) and a perturbed version of the dataset to simulate distribution shifts. The results demonstrate the effectiveness of the MLM in assessing the reliability of predictions and highlight its potential in enhancing the interpretability and risk mitigation capabilities of neural networks. The implications of this work extend beyond image classification, with ongoing applications in autonomous systems, such as self-driving cars, to improve the safety and reliability of decision-making in complex, real-world environments.en_US
dc.description.sectionheadersMachine Learning and LLM-enabled Visual Analytics
dc.description.seriesinformationComputer Graphics and Visual Computing (CGVC)
dc.identifier.doi10.2312/cgvc.20241239
dc.identifier.isbn978-3-03868-249-3
dc.identifier.pages9 pages
dc.identifier.urihttps://doi.org/10.2312/cgvc.20241239
dc.identifier.urihttps://diglib.eg.org/handle/10.2312/cgvc20241239
dc.publisherThe Eurographics Associationen_US
dc.rightsAttribution 4.0 International License
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subjectCCS Concepts: Computing methodologies → Machine learning; Computer vision
dc.subjectComputing methodologies → Machine learning
dc.subjectComputer vision
dc.titleThe Misclassification Likelihood Matrix: Some Classes Are More Likely To Be Misclassified Than Othersen_US
Files
Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
cgvc20241239.pdf
Size:
603.07 KB
Format:
Adobe Portable Document Format
Download
Collections
Computer Graphics & Visual Computing (CGVC) 2024