Overcoming Common Flaws in the Evaluation of Selective Classification Systems
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2024
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866914979799629824 |
|---|---|
| author | Traub, Jeremias Bungert, Till J. Lüth, Carsten T. Baumgartner, Michael Maier-Hein, Klaus H. Maier-Hein, Lena Jaeger, Paul F |
| author_facet | Traub, Jeremias Bungert, Till J. Lüth, Carsten T. Baumgartner, Michael Maier-Hein, Klaus H. Maier-Hein, Lena Jaeger, Paul F |
| contents | Selective Classification, wherein models can reject low-confidence predictions, promises reliable translation of machine-learning based classification systems to real-world scenarios such as clinical diagnostics. While current evaluation of these systems typically assumes fixed working points based on pre-defined rejection thresholds, methodological progress requires benchmarking the general performance of systems akin to the $\mathrm{AUROC}$ in standard classification. In this work, we define 5 requirements for multi-threshold metrics in selective classification regarding task alignment, interpretability, and flexibility, and show how current approaches fail to meet them. We propose the Area under the Generalized Risk Coverage curve ($\mathrm{AUGRC}$), which meets all requirements and can be directly interpreted as the average risk of undetected failures. We empirically demonstrate the relevance of $\mathrm{AUGRC}$ on a comprehensive benchmark spanning 6 data sets and 13 confidence scoring functions. We find that the proposed metric substantially changes metric rankings on 5 out of the 6 data sets. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2407_01032 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Overcoming Common Flaws in the Evaluation of Selective Classification Systems Traub, Jeremias Bungert, Till J. Lüth, Carsten T. Baumgartner, Michael Maier-Hein, Klaus H. Maier-Hein, Lena Jaeger, Paul F Machine Learning Computer Vision and Pattern Recognition Methodology Selective Classification, wherein models can reject low-confidence predictions, promises reliable translation of machine-learning based classification systems to real-world scenarios such as clinical diagnostics. While current evaluation of these systems typically assumes fixed working points based on pre-defined rejection thresholds, methodological progress requires benchmarking the general performance of systems akin to the $\mathrm{AUROC}$ in standard classification. In this work, we define 5 requirements for multi-threshold metrics in selective classification regarding task alignment, interpretability, and flexibility, and show how current approaches fail to meet them. We propose the Area under the Generalized Risk Coverage curve ($\mathrm{AUGRC}$), which meets all requirements and can be directly interpreted as the average risk of undetected failures. We empirically demonstrate the relevance of $\mathrm{AUGRC}$ on a comprehensive benchmark spanning 6 data sets and 13 confidence scoring functions. We find that the proposed metric substantially changes metric rankings on 5 out of the 6 data sets. |
| title | Overcoming Common Flaws in the Evaluation of Selective Classification Systems |
| topic | Machine Learning Computer Vision and Pattern Recognition Methodology |
| url | https://arxiv.org/abs/2407.01032 |