Evaluating the Quality of Machine Learning Explanations: A Survey on Methods and Metrics
Open accessThe DOI matches its Crossref record; 88 of the 88 references it cites were found in a registry or an index.
Abstract
The most successful Machine Learning (ML) systems remain complex black boxes to end-users, and even experts are often unable to understand the rationale behind their decisions. The lack of transparency of such systems can have severe consequences or poor uses of limited valuable resources in medical diagnosis, financial decision-making, and in other high-stake domains. Therefore, the issue of ML explanation has experienced a surge in interest from the research community to application domains. While numerous explanation methods have been explored, there is a need for evaluations to quantify the quality of explanation methods to determine whether and to what extent the offered explainability achieves the defined objective, and compare available explanation methods and suggest the best explanation from the comparison for a specific task. This survey paper presents a comprehensive overview of methods proposed in the current literature for the evaluation of ML explanations. We identify properties of explainability from the review of definitions of explainability. The identified properties of explainability are used as objectives that evaluation metrics should achieve. The survey found that the quantitative metrics for both model-based and example-based explanations are primarily used to evaluate the parsimony/simplicity of interpretability, while the quantitative metrics for attribution-based explanations are primarily used to evaluate the soundness of fidelity of explainability. The survey also demonstrated that subjective measures, such as trust and confidence, have been embraced as the focal point for the human-centered evaluation of explainable systems. The paper concludes that the evaluation of ML explanations is a multidisciplinary research topic. It is also not possible to define an implementation of evaluation metrics, which can be applied to all explanation methods.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
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 cited source says what the claim attributes to it.
In the citing paper
“They found that some properties (clarity, broadness, and completeness) are still in shortage of appropriate metrics, and so is the class of explanations that are example-based.”
From Current Challenges and Future Opportunities for XAI in Machine Learning-Based Clinical Decision Support Systems: A Systematic Review (2021) · cites this paper as [141]
In the citing paper
“It is not possible to come up with quantitative metrics that are applicable for all interpretability methods [169].”
From Transparency of deep neural networks for medical image analysis: A review of interpretability methods (2022) · cites this paper as [169]
In the citing paper
“They conclude their survey by stating that”
From Current Challenges and Future Opportunities for XAI in Machine Learning-Based Clinical Decision Support Systems: A Systematic Review (2021) · cites this paper as [141]
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
A Unified Approach to Interpreting Model Predictions
Lundberg, S., & Lee, S.-I. (2017) · arXiv · doi:10.48550/arxiv.1705.07874 · Cited by 23,707 · cited 2×
“One way to objectively evaluate explanations is to verify whether the explanation mechanism satisfies (or does not satisfy) certain axioms, or properties [34].”
Background"Why Should I Trust You?": Explaining the Predictions of Any Classifier
Ribeiro, M. T., Singh, S., & Guestrin, C. (2016) · Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining · doi:10.1145/2939672.2939778 · Cited by 17,555
Distilling the Knowledge in a Neural Network
Hinton, G., Vinyals, O., & Dean, J. (2015) · arXiv · doi:10.48550/arxiv.1503.02531 · Cited by 14,204 · cited 1×
“[39] explained an ensemble of models, such as neural networks with a single model effectively.”
BackgroundCausality: Models, Reasoning, and Inference
Pearl, J. (2009) · doi:10.1017/cbo9780511803161 · Cited by 11,653
Rudin, C. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-019-0048-x · Cited by 10,561 · cited 1×
“Explanation is also a domain-specific note, and there cannot be an all-purpose explanation [22], and different types of explanations might be useful.”
BackgroundPeeking Inside the Black-Box: A Survey on Explainable Artificial Intelligence (XAI)
Adadi, A., & Berrada, M. (2018) · IEEE Access · doi:10.1109/access.2018.2870052 · Cited by 6,294 · cited 2×
“The machine learning explainability has been reviewed thoroughly in recent years [10,14,17,23,28,31,32].”
BackgroundLipton, Z. C. (2018) · Queue · doi:10.1145/3236386.3241340 · Cited by 5,427
Explanation in artificial intelligence: Insights from the social sciences
Miller, T. (2019) · Artificial Intelligence · doi:10.1016/j.artint.2018.07.007 · Cited by 5,302
Axiomatic attribution for deep networks
Sundararajan, M., Taly, A., & Yan, Q. (2017) · arXiv · doi:10.48550/arxiv.1703.01365 · Cited by 4,087 · cited 4×
“…Main effect complexity [46] Level of (dis)agreement [82] Attribution-based explanations Monotonicity [80] Non-sensitivity [80,83], Sensitivity [84,85] Effective complexity [80] Remove and retrain [86] Recall of important features [33] Implementation invariance [85] Selectivity [87] Continuity…”
BackgroundInfluentialTowards A Rigorous Science of Interpretable Machine Learning
Doshi-Velez, F., & Kim, B. (2017) · arXiv · doi:10.48550/arxiv.1702.08608 · Cited by 3,121 · cited 4×
“The advantages of application-grounded and human-grounded evaluations are that they can provide direct and strong evidence of success of explanations [9].”
BackgroundMethodInfluential
Research works printed but not linked: 5
- Zhou, J., and Chen, F. (2018). Human and Machine Learning: Visible, Explainable, Trustworthy and Transparent, Springer.
- Lage, I., and Doshi-Velez, F. (2020, January 17). Human-in-the-Loop Learning of Interpretable and Intuitive Representations. Proceedings of the ICML Workshop on Human Interpretability in Machine Learning, Vienna, Austria.
- Becker, B., Kohavi, R., and Sommerfield, D. (2001). Visualizing the Simple Bayesian Classifier. Information Visualization in Data Mining and Knowledge Discovery, Morgan Kaufmann Publishers Inc.
- Poursabzi-Sangdeh, F., Goldstein, D.G., Hofman, J.M., Vaughan, J.W., and Wallach, H. (2019). Manipulating and Measuring Model Interpretability. arXiv.
- Yang, M., and Kim, B. (2019). Benchmarking Attribution Methods with Relative Feature Importance. arXiv, 1907.
Cited by
Explainable AI (XAI): A systematic meta-survey of current challenges and future opportunities
Saeed, W., & Omlin, C. (2023) · Knowledge-Based Systems · doi:10.1016/j.knosys.2023.110273 · Cited by 783
Antoniadi, A. M., Du, Y., Guendouz, Y., Wei, L., Mazo, C., Becker, B. A., & Mooney, C. (2021) · Applied Sciences · doi:10.3390/app11115088 · Cited by 629
“A summary of many quantitative metrics for the evaluation of explainability properties (i.e., clarity, broadness, parsimony, completeness, and soundness) for different explanation types, is presented in the work of Zhou et al. [141].”
BackgroundLongo, L., Brcic, M., Cabitza, F., Choi, J., Confalonieri, R., Ser, J. D., Guidotti, R., Hayashi, Y., Herrera, F., Holzinger, A., et al. (2024) · Information Fusion · doi:10.1016/j.inffus.2024.102301 · Cited by 626
“Furthermore, XAI has evolved into an exceedingly multidisciplinary, interdisciplinary, and transdisciplinary field.”
BackgroundSelf-citationNauta, M., Trienes, J., Pathak, S., Nguyen, E., Peters, M., Schmitt, Y., Schlötterer, J., van Keulen, M., & Seifert, C. (2023) · ACM Computing Surveys · doi:10.1145/3583558 · Cited by 590
“[314] that presents a concise overview and discussion regarding evaluating XAI without a structured literature review, and the recent work of Vilone and Longo [263] that present a list of concepts related to explainability and discuss evaluation methods in combination with explainable AI methods.”
BackgroundInfluentialSalahuddin, Z., Woodruff, H. C., Chatterjee, A., & Lambin, P. (2022) · Computers in Biology and Medicine · doi:10.1016/j.compbiomed.2021.105111 · Cited by 508
Metadata
- Authors
- Zhou, J., Gandomi, A. H., Chen, F., & Holzinger, A.
- Affiliations
- University of Technology Sydney (AU)
- Medical University of Graz (AT)
- Corresponding author
- Zhou, Jianlong
- Gandomi, Amir H.
- Venue
- Electronics
- Publisher
- MDPI AG
- Published
- 2021-03-04
- Volume
- 10
- Issue
- 5
- Pages
- 593
- ISSN
- 2079-9292 (ISSN-L)
- OpenAlex
- W3133543405
- Cited by
- 606
- References
- 80
- Citations by year
- 2026: 75 · 2025: 138 · 2024: 179 · 2023: 109 · 2022: 84 · 2021: 20
- FWCI
- 34.65
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- Electronics (publishedVersion, cc-by)
- OPUS - Open Publications of UTS Scholars (University of Technology Sydney) (submittedVersion, other-oa)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Funders
- University of Technology Sydney
- Austrian Science Fund (P 32554)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Physical Sciences › Computer Science › Software › Software Testing and Debugging Techniques
- Physical Sciences › Computer Science › Artificial Intelligence › Adversarial Robustness in Machine Learning
- Keywords
- explainable machine learning, explanation quality evaluation, evaluation metrics, explainability, model-based explanation, example-based explanations, interpretability, fidelity, human-centered evaluation, black-box models
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
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.