CoChat
Sign Up
Cited by 7Adversarial Robustness in Machine LearningSafety Systems Engineering in AutonomyImbalanced Data Classification Techniques

Bringing Machine Learning Classifiers Into Critical Cyber-Physical Systems: A Matter of Design

Open access
Sayin, Burcu, Zoppi, Tommaso, Marchini, Nicolò, Khokhar, Fahad Ahmed, Passerini, Andrea·2025·IEEE Access
DOI 10.1109/access.2025.3568501
Open this paper in CoChat — free account

Abstract

Domain experts are striving to solve classification problems through the application of Machine Learning (ML) models, aiming for the highest accuracy. However, classifiers are inherently prone to misclassifications, especially when encountering unknown input data. This complicates their deployment in critical Cyber-Physical Systems (CPSs), where misclassifications can have detrimental effects on people, the environment, or infrastructure. This paper argues that ML classifiers should not be designed or evaluated in isolation but rather conceptualized as Critical System Classifiers (CSCs), which can reject uncertain predictions and trigger mitigation strategies in accordance to the encompassing CPS. We present an high-level architecture for CSCs that supports black-box integration, preprocessing for unknown detection, post-hoc calibration, and cost-sensitive thresholding, emphasizing the importance of evaluating classifiers using cost-aware metrics that account for rejections. We validate our proposal through experiments on tabular datasets related to failure prediction, intrusion and error detection, which are common applications of classifiers within CPSs. Key findings are: i) cost-sensitive evaluation promotes the usage of different classifiers compared to standard performance metrics, ii) tree-based models outperform statistical ones, classification-wise, iii) prediction calibration and rejection strategies provide a robust notion of confidence, and iv) combining multiple uncertainty-based rejectors achieves a favorable balance between high accuracy, low rejection rate, and cost. The analyses above were conducted using our publicly available CINNABAR GitHub repository for experimentations with CSCs. Overall, our study offers a system-level perspective, software architecture and implementation on classifier deployment in critical CPSs domains, paving the way for a trustworthy ML integration into real-world infrastructures.

Integrity signals

Registry

Crossref · 10.1109/access.2025.3568501

Cites

76 of 76 references verified0 citation errors · 0 not checkable

Statistics

Recomputing p-values needs the paper’s full text.

Claims the paper makes

We couldn’t retrieve this paper’s full text, so its claims and self-check weren’t run. The registry and citation checks still apply.

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • The Critical System Classifier (CSC) framework highlights that uncertainty-aware calibration and rejection (abstention) are essential in safety-critical CPS, where confidently misclassifying unknown or out-of-distribution traffic can be more harmful than refusing to decide.

    “the Critical System Classifier (CSC) framework highlights that uncertainty-aware calibration and rejection (abstention) are essential in safety-critical CPS, where confidently misclassifying unknown or out-of-distribution traffic can be more harmful than refusing to decide [56].”

    From AI-Based Anomaly Detection in Industrial Control and Cyber–Physical Systems: A Data-Type-Oriented Systematic Review (2025) · cites this paper as [56]

  • CSC explicitly rejects uncertain predictions and triggers system-level mitigation strategies rather than maximizing accuracy.

    “Rather than 'maximizing accuracy,' CSC explicitly rejects uncertain predictions and triggers system-level mitigation strategies.”

    From AI-Based Anomaly Detection in Industrial Control and Cyber–Physical Systems: A Data-Type-Oriented Systematic Review (2025) · cites this paper as [56]

  • Experiments in Sayin et al. show that tree-based ensembles outperform statistical models, and combining calibration with rejection yields an effective balance between accuracy, acceptable rejection rates, and reduced risk cost.

    “Experiments show that tree-based ensembles outperform statistical models, and combining calibration with rejection yields an effective balance between accuracy, acceptable rejection rates, and reduced risk cost.”

    From AI-Based Anomaly Detection in Industrial Control and Cyber–Physical Systems: A Data-Type-Oriented Systematic Review (2025) · cites this paper as [56]

  • Sayin et al. [56] proposed the concept of a Critical System Classifier (CSC) for deploying machine-learning classifiers in safety-critical cyber-physical systems such as autonomous vehicles, medical devices, and ICS.

    “Sayin et al. [56] proposed the concept of a Critical System Classifier (CSC) for deploying machine-learning classifiers in safety-critical cyber-physical systems such as autonomous vehicles, medical devices, and ICS.”

    From AI-Based Anomaly Detection in Industrial Control and Cyber–Physical Systems: A Data-Type-Oriented Systematic Review (2025) · cites this paper as [56]

Cites

Flagged first, then by citations

1 of the 77 references couldn’t be matched to a known paper.

Cited by

Sorted by citations

Related

Authors
Sayin, B., Zoppi, T., Marchini, N., Khokhar, F. A., & Passerini, A.
Affiliations
  • University of Trento (IT)
  • University of Florence (IT)
ORCID
Venue
IEEE Access
Publisher
Institute of Electrical and Electronics Engineers (IEEE)
Published
2025
Volume
13
Pages
94858-94877
ISSN
2169-3536 (ISSN-L)
DOI
10.1109/access.2025.3568501
OpenAlex
W4410226707
Cited by
7
References
77
Citations by year
2026: 4 · 2025: 3
FWCI
2.83
Type
journal article
Language
en
Open access
Licence
Locations
  • IEEE Access (publishedVersion, cc-by-nc-nd)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • Florence Research (University of Florence) (submittedVersion, cc-by-nc-nd)
  • Iris (University of Trento) (submittedVersion, cc-by-nc-nd)
Funders
  • Progetti di Ricerca di Interesse Nazionale (PRIN) 2022 Project S2 (202297YF75)
  • Project SEcurity and RIghts in the CyberSpace (SERICS) under the Ministero dell’Università e Ricerca (MUR) National Recovery and Resilience Plan funded by European Union–NextGenerationEU (PE00000014;)
  • Cognitive Safety with Point Clouds (CogniSafe3D) Eurostars 3, Call 6 by European Union.
  • Horizon Europe Project TANGO funded by European Union (101120763)
Topics
  • Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
  • Physical Sciences › Engineering › Safety, Risk, Reliability and Quality › Safety Systems Engineering in Autonomy
  • Physical Sciences › Computer Science › Artificial Intelligence › Imbalanced Data Classification Techniques
Keywords
machine learning classifiers, forecast calibration, failure prediction, intrusion detection, error detection

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review