Trusting My Predictions: On the Value of Instance-Level Analysis
ACM Computing Surveys · 31.07.2024 · 56 (7) · 1-28
DOI: 10.1145/3615354 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Computing Surveys, online erschienen am 09.04.2024. Verfasst von Ana C. Lorena und 2 weiteren Autor:innen. Die Arbeit wurde bisher 30-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: Brazilian Research Agencies FAPESP, CNPq.
Abstract (Originalsprache)
Machine Learning solutions have spread along many domains, including critical applications. The development of such models usually relies on a dataset containing labeled data. This dataset is then split into training and test sets and the accuracy of the models in replicating the test labels is assessed. This process is often iterated in a cross-validation procedure for obtaining average performance estimates. But is the average of the predictive performance on test sets enough for assessing the trustfulness of a Machine Learning model? This paper discusses the importance of knowing which individual observations of a dataset are more challenging than others and how this characteristic can be measured and used in order to improve classification performance and trustfulness. A set of strategies for measuring the hardness level of the instances of a dataset is surveyed and a Python package containing their implementation is provided.
Förderung
- Brazilian Research Agencies FAPESP
- CNPq
Zitieren
APA
Lorena, A. C., Paiva, P. Y. A., & Prudêncio, R. B. C. (2024). Trusting My Predictions: On the Value of Instance-Level Analysis. ACM Computing Surveys, 56(7), 1-28. https://doi.org/10.1145/3615354
Vancouver
Lorena AC, Paiva PYA, Prudêncio RBC. Trusting My Predictions: On the Value of Instance-Level Analysis. ACM Computing Surveys. 2024;56(7):1-28. doi:10.1145/3615354
BibTeX
@article{Lorena2024,
author = {Lorena, Ana C. and Paiva, Pedro Y. A. and Prudêncio, Ricardo B. C.},
title = {Trusting My Predictions: On the Value of Instance-Level Analysis},
journal = {ACM Computing Surveys},
year = {2024},
volume = {56},
number = {7},
pages = {1--28},
doi = {10.1145/3615354}
}
EUVIMED