Herd, Benjamin and Burton, Simon orcid.org/0000-0001-9040-8752 (2024) Can you trust your ML metrics? Using Subjective Logic to determine the true contribution of ML metrics for safety. In: 39th Annual ACM Symposium on Applied Computing, SAC 2024. 39th Annual ACM Symposium on Applied Computing, SAC 2024, 08-12 Apr 2024 Proceedings of the ACM Symposium on Applied Computing . Association for Computing Machinery, Inc , ESP , pp. 1579-1586.
Abstract
Metrics such as accuracy, precision, recall, F1 score, etc. are generally used to assess the performance of machine learning (ML) models. From a safety perspective, relying on such single point estimates to evaluate safety requirements is problematic since they only provide a partial and indirect evaluation of the true safety risk associated with the model and its potential errors. In order to obtain a better understanding of the performance insufficiencies in the model, factors that could influence the quantitative evaluation of safety requirements such as test sample size, dataset size and model calibration need to be taken into account. In safety assurance, arguments typically combine complementary and diverse evidence to strengthen confidence in the safety claims. In this paper, we make a first step towards a more formal treatment of uncertainty in ML metrics by proposing a framework based on Subjective Logic that allows for modelling the relationship between primary and secondary pieces of evidence and the quantification of resulting uncertainty. Based on experiments, we show that single point estimates for common ML metrics tend to overestimate model performance and that a probabilistic treatment using the proposed framework can help to evaluate the probable bounds of the actual performance.
Metadata
Item Type: | Proceedings Paper |
---|---|
Authors/Creators: |
|
Copyright, Publisher and Additional Information: | Publisher Copyright: © 2024 Copyright held by the owner/author(s). |
Keywords: | machine learning,safety assurance,subjective logic,uncertainty |
Dates: |
|
Institution: | The University of York |
Academic Units: | The University of York > Faculty of Sciences (York) > Computer Science (York) |
Depositing User: | Pure (York) |
Date Deposited: | 06 Aug 2025 09:40 |
Last Modified: | 06 Aug 2025 09:40 |
Published Version: | https://doi.org/10.1145/3605098.3635966 |
Status: | Published |
Publisher: | Association for Computing Machinery, Inc |
Series Name: | Proceedings of the ACM Symposium on Applied Computing |
Identification Number: | 10.1145/3605098.3635966 |
Related URLs: | |
Open Archives Initiative ID (OAI ID): | oai:eprints.whiterose.ac.uk:230126 |
Download
Filename: 3605098.3635966.pdf
Description: Can you trust your ML metrics? Using Subjective Logic to determine the true contribution of ML metrics for safety
Licence: CC-BY 2.5