Robustness under noise: assessing the impact of perturbed key attributes on machine learning models
Open Access
- 23 February 2026
- journal article
- research article
- Published by Springer Nature in International Journal of Data Science and Analytics
- Vol. 22 (1) , 71
Abstract
Machine learning models often struggle when confronted with the complexities of real-world data, such as noise and incomplete information, which pose significant challenges for practical applications. This paper presents a systematic framework for assessing the robustness of machine learning models against these adversities. In particular, we investigate how noise affecting the most critical attribute of a dataset impacts on model predictions. Our methodology uses a variety of datasets, a wide range of machine learning models, and a customizable noise induction function, allowing for comprehensive evaluations tailored to specific application needs. Through an extensive experimental analysis, we identify the vulnerabilities of models when their critical attributes are perturbed. In addition, by applying hierarchical clustering, we categorise machine learning families based on their robustness, establishing a taxonomy that highlights the susceptibility of different models to noise. Our findings provide important insights into the interplay between model performance and noise levels and improve the understanding of model behaviour under adverse conditions.Keywords
Funding Information
- Ministerio de Ciencia, Innovación y Universidades (ID2024-162030OB-I00)
- Universitat Politècnica de València (PAID-01-22)
- Generalitat Valenciana (CIPROM/2022/6)
- Cátedra ENIA-UPV (TSI-100930-2023-9)
This publication has 58 references indexed in Scilit:
- DMGCL: Denoising Multi-view Graph Contrastive Learning for Robust Recommendation2025
- An assessment framework for explainable AI with applications to cybersecurityArtificial Intelligence Review, 2025
- NPGCL: neighbor enhancement and embedding perturbation with graph contrastive learning for recommendationApplied Intelligence, 2025
- Noise Tolerance and Robustness Ranking in Machine Learning Models2024
- Safety-critical computer vision: an empirical survey of adversarial evasion attacks and defenses on computer vision systemsArtificial Intelligence Review, 2023
- Comprehensive survey on hierarchical clustering algorithms and the recent developmentsArtificial Intelligence Review, 2022
- Reluplex: An Efficient SMT Solver for Verifying Deep Neural Networks2017
- Analyzing the presence of noise in multi-class problems: alleviating its influence with the One-vs-One decompositionKnowledge and Information Systems, 2012
- Class Noise vs. Attribute Noise: A Quantitative StudyArtificial Intelligence Review, 2004
- KNN Model-Based Approach in Classification2003