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
Items 401–500 of 676
Shevskaya, N. V., Shishkin, I. V., Stepanov, V. D., & Oblizanov, A. D. (2022) · International Conference on Neural Networks and Neurotechnologies (NeuroNT) · doi:10.1109/neuront55429.2022.9805526 · Cited by 3
Szczepański, M., Pawlicki, M., Kozik, R., & Choraś, M. (2022) · IEEE International Conference on Data Science and Advanced Analytics (DSAA) · doi:10.1109/dsaa54385.2022.10032372 · Cited by 3
“Selection of the adequate explainability metric is not a trivial task, often limited by the class of the explanation method [42].”
BackgroundExplainable AI for CPS-Based Manufacturing Workcell
Soon, R. J., Sang, D. V., Chng, C.-B., & Chui, C. (2023) · International Conference on System Science and Engineering · doi:10.1109/icsse58758.2023.10227195 · Cited by 3
“This is substantiated in [28], [30], [38], where it was illustrated that the use of rule-based analyses did not necessarily increase trust in the system’s outputs.”
BackgroundMethodResultData Engineering Challenges in AI automation
Reddy, D. (2023) · International Conference on Computing, Electronics & Communications Engineering (iCCECE) · doi:10.1109/iccece59400.2023.10238496 · Cited by 3
“A notable tool in this category is Google’s AI Platform, which provides capabilities for managing machine learning workflows, including model versioning, tracking, and deployment [71].”
MethodGender Disparities in Child Custody Sentencing in Spain: a Data Driven Analysis
Riera, J., Solans, D., Karimi-Haghighi, M., Castillo, C., & Calsamiglia, C. (2023) · International Conference on Artificial Intelligence and Law · doi:10.1145/3594536.3595135 · Cited by 4
Explainability of Artificial Intelligence Systems: A Survey
Hossain, M., Das, S., Krishnamurthy, B., & Shiva, S. G. (2023) · International Symposium on Networks, Computers and Communications (ISNCC) · doi:10.1109/isncc58260.2023.10323642 · Cited by 4
“…fundamental guidelines that must be followed to maintain explanation qualities like fidelity (reasonable explanation of the inner system working), understandability, sufficiency, lower additional computation overhead, and efficiency (not affecting the computational speed of the ML system) [7]–[9].”
BackgroundCOSE: A Consistency-Sensitivity Metric for Saliency on Image Classification
Daroya, R., Sun, A., & Maji, S. (2023) · IEEE International Conference on Computer Vision Workshops · doi:10.1109/iccvw60793.2023.00022 · Cited by 5
“Quantitative methods can help analyze other factors and narrow down methods to examine more closely [33].”
BackgroundReal Time Quality Assessment in a Production Line: Machine Learning Approach
Mahapatra, R., & Gaurav, K. (2023) · International Conference on Modeling, Simulation & Intelligent Computing (MoSICom) · doi:10.1109/mosicom59118.2023.10458752 · Cited by 3
“For the evaluation metrics on the test set, refer to Table 2, and the confusion matrix is shown in Figure 9 [37].”
MethodQu, J., Arguello, J., & Wang, Y. (2024) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3630106.3658547 · Cited by 5
Explaining poor performance of text-based machine learning models for vulnerability detection
Napier, K., Bhowmik, T., & Chen, Z. (2024) · Empirical Software Engineering · doi:10.1007/s10664-024-10519-8 · Cited by 3
“Some aspects of a model’s behavior, particularly those of highly complex models or those operating in dynamic environments, may not be captured at all in the explanations Zhou et al. (2021).”
The modern methods of data preservation in libraries. The key aspects and solutions
Beskaravainaya, E. V., & Митрошин, И. А. (2024) · Scientific and Technical Libraries · doi:10.33186/1027-3689-2024-8-78-97 · Cited by 3
Evaluation Methodology for Interpretation Methods of Predictive Quality Models
Schulze, T., Buschmann, D., & Schmitt, R. H. (2024) · Procedia CIRP · doi:10.1016/j.procir.2024.08.366 · Cited by 3
Personalized Guidance for Moroccan Students: An Approach Based on Machine Learning and Big Data
Badrani, M., Marouan, A., Kannouf, N., & Chetouani, A. (2025) · International Journal of Engineering Pedagogy (iJEP) · doi:10.3991/ijep.v15i1.51985 · Cited by 6
Hendriks, D., Spitzer, P., Kühl, N., & Satzger, G. (2026) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2026.3671872 · Cited by 6
“Based on literature on explainability [82], [83], [84], [85], we defined explanation quality through five dimensions: (1) Plausibility is the characteristic of an explanation to be sound and reasonable [86].”
GANExplainer: GAN-based Graph Neural Networks ExplainerPreprint
Li, Y., Zhou, J. J., Zheng, B., & Chen, F. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2301.00012 · Cited by 3
“After comparing the characteristic of each metric [11,28], we chose quantitative and qualitative evaluation.”
MethodSelf-citationMotif Interactions Affect Post-Hoc Interpretability of Genomic Convolutional Neural NetworksPreprint
Lemanczyk, M., Bartoszewicz, J. M., & Renard, B. Y. (2024) · bioRxiv (Cold Spring Harbor Laboratory) · doi:10.1101/2024.02.15.580353 · Cited by 4
“The quality of contribution scores can be affected by multiple factors besides model complexity so the evaluation of interpretability performance for machine learning models is crucial [12, 13, 14].”
BackgroundRagodos, R., Wang, T., Feng, L., & Hu, Y. J. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4915307 · Cited by 5
Symbolic approaches for explainable artificial intelligence
Bellucci, M. (2023) · theses.fr (ABES) · doi:10.70675/862d4140z032fz4a15zb1ffz1d6d85c7783e · Cited by 7
Javeed, A., Anderberg, P., Saleem, M. A., Ghazi, A. N., & Berglund, J. S. (2024) · International Journal of Imaging Systems and Technology · doi:10.1002/ima.23221 · Cited by 4
“The holdout validation method is often used in the literature as a standard to evaluate the performance of ML-based diagnostic systems [45–47].”
Shiri, M., Moharramnejad, S., Estakhr, A., Fareghi, S., Najafinezhad, H., Khorasani, S. K., Afarinesh, A., & Eshraghi-Nejad, M. (2025) · PLoS ONE · doi:10.1371/journal.pone.0325454 · Cited by 2
“This measure can be compared directly to Zhou et al. [26] method that defines a stable genotype as having low variance in the effects of GEI [ var ( ˆ ge ) ].”
Riding the Paradox: How AI Consultants Manage the Tradeoff Between Explainability and Performance
Benyakhlaf, M., Lamkadem, A., & Okoli, C. (2025) · Information Systems Management · doi:10.1080/10580530.2025.2506369 · Cited by 2
“In addition, respondents shared that some precise indicators for evaluating an explanation permit the assessment of the coherence of the model whose results are already known for a given dataset (Zhou et al., 2021).”
InfluentialTocchetti, A., Bianchi, M., Campi, R., Santis, A. D., & Brambilla, M. (2025) · Elsevier eBooks · doi:10.1016/b978-0-44-340553-2.00012-5 · Cited by 2
Guo, D. (2025) · International Conference on Intelligent Computing and Knowledge Extraction (ICICKE) · doi:10.1109/icicke65317.2025.11136300 · Cited by 2
“R ELATED W ORKS Zhou J. et al. (2021) [15] conducted a comprehensive survey on evaluating the quality of machine learning explanations by examining various interpretability methods and corresponding metrics.”
Knowing What You Cannot Explain: Learning to Reject Low-Quality Explanations
Stradiotti, L., Pesenti, D., Teso, S., & Davis, J. J. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-37657-2_11 · Cited by 4
EmoVec: emotion detection in Malayalam using custom word embeddings
Anuja, K., Raj, P. C. R., & Babu, K. R. R. (2025) · Language Resources and Evaluation · doi:10.1007/s10579-025-09892-7 · Cited by 2
“The trained models are tested against test samples, and the predictions are evaluated using evaluation parameters such as precision (P), recall (R), F1 score (F1), and accuracy (A) (Hossin & Sulaiman, 2015; Liu et al., 2014; Zhou et al., 2021).”
An Evaluation of XAI Methods applied to Information Disorder Detection Models
Gruensteidl, M. N., & Kirrane, S. (2026) · Knowledge-Based Systems · doi:10.1016/j.knosys.2026.115305 · Cited by 2
Alarcon, G. M., Jessup, S. A., Meyers, S. K., Willis, S. M., Harris, K. N., Johnson, D., Noblick, J., & Bennette, W. D. (2026) · IEEE Transactions on Human-Machine Systems · doi:10.1109/thms.2026.3651060 · Cited by 2
Nasution, M. H. I., Habib, S., Ehsan, F., Gulzar, M. M., Al-Awami, A. T., & Khan, B. (2026) · Results in Engineering · doi:10.1016/j.rineng.2026.111055 · Cited by 2
A Review of Taxonomies of Explainable AI (XAI) Evaluation
Speith, T. (2026) · ACM Conference on Fairness, Accountability, and Transparency (FAccT) · doi:10.1145/3805689.3806537 · Cited by 3
Explainable AI for Deep Visual Recognition: Evaluation, Methods, and Open Challenges
Alharbi, K. N. (2026) · Electronics · doi:10.3390/electronics15143222 · Cited by 2
Human-AI Interfaces are a Central Component of Trustworthy AI
Plass, M., Kargl, M., Evans, T., Brčić, L., Regitnig, P., Geißler, C., Carvalho, R., Jansen, C., Zerbe, N., Holzinger, A., et al. (2022) · Intelligent systems reference library · doi:10.1007/978-3-031-12807-3_11 · Cited by 6
Self-citationiBMSR: Intelligent Body Mass Status Recognition from Respiratory Auscultation
Adhikary, S., Ghosh, A., & Nandi, S. (2023) · International Conference on Communication Systems & Networks (COMSNETS) · doi:10.1109/comsnets56262.2023.10041309 · Cited by 2
“Finally, the performances are measured by accuracy, precision, recall, F1 score, Jaccard Score, Kohen Kappa statistics, Hamming Loss, Mathews Correlation Coefficient, training and testing time [19].”
MethodDeng, Y., Weng, Z., & Zhang, T. (2023) · Journal of Multimedia Information System · doi:10.33851/jmis.2023.10.2.199 · Cited by 2
GX-HUI: Global Explanations of AI Models based on High-Utility Itemsets
Napolitano, D., & Cagliero, L. (2023) · IEEE Annual International Computer Software and Applications Conference (COMPSAC) · doi:10.1109/compsac57700.2023.00045 · Cited by 2
Quantification of Explainability in Black Box Models using Complexity Measures
Nneke, N., Lloyd, H., & Crockett, K. A. (2023) · International Conference on Innovations in Information Technology · doi:10.1109/iit59782.2023.10366490 · Cited by 2
“No standardized objective metrics to evaluate explanation methods [13] currently exist.”
BackgroundMethodResultEvaluation and Validation of Design Metrics in Machine Learning for Depression Detection
Ajmal, S., Usman, M., & Javed, A. (2023) · International Conference on Emerging Technologies · doi:10.1109/icet59753.2023.10375035 · Cited by 2
“Likewise, [20] designed 3 metrics on the testing level but it focused on explainability to test ML explanations.”
BackgroundExploring Multi-Task Learning for Explainability
Charalampakos, F., & Koutsopoulos, I. (2024) · Communications in computer and information science · doi:10.1007/978-3-031-50396-2_20 · Cited by 4
“Several metrics have been developed to objectively assess the quality of explanations according to different criteria [3].”
BackgroundLai, X.-W., Zhang, K., Li, Z.-X., Zheng, Q., & Ding, G. (2023) · 2023 IEEE Smart World Congress (SWC) · doi:10.1109/swc57546.2023.10448633 · Cited by 3
“Zhou et al.[8] presented a survey on methods and metrics of machine learning explanations.”
MethodBalasundaram, R. J., Karthick, G., Shah, P. B., Vellore-Nagarajan, D., Arivazhagan, R., Earnest, P., & Dutta, S. (2024) · MIGRATION LETTERS · doi:10.59670/ml.v21is7.8919 · Cited by 2
“Various metrics are employed to assess their performance quantitatively in evaluating sentiment analysis models (Zhou, 2021).”
MethodAn AutoML-based approach for automatic traffic incident detection in smart cities
Γκιόκα, Γ., Domínguez, M., & Mentzas, G. N. (2024) · Intelligent Decision Technologies · doi:10.3233/idt-240231 · Cited by 2
“These standard performance metrics have been widely utilized in the Machine Learning domain due to their effectiveness in assessing algorithm performance [59].”
Nizam, T., Zafar, S., Biswas, S. S., & Hussain, I. (2024) · Lecture notes in networks and systems · doi:10.1007/978-981-97-1260-1_13 · Cited by 2
Nguyen, M., Truong, H., & Truong-Huu, T. (2024) · IEEE/ACM International Conference on AI Engineering – Software Engineering for AI (CAIN) · doi:10.1145/3644815.3644964 · Cited by 2
Transparency and trustworthiness: Exploring humanmachine interaction in an image classification task
Meyers, S., Capiola, A., Alarcon, G. M., & Bennette, W. D. (2024) · IEEE International Conference on Human-Machine Systems (ICHMS) · doi:10.1109/ichms59971.2024.10555635 · Cited by 2
“XAI – like ML – is a subfield within AI that seeks to increase the transparency of a model for users [12].”
BackgroundOn Quantification of the Nonlinearity of PPV in Model Evaluation with Imbalanced Datasets
Zhu, Q. (2024) · International Journal of Pattern Recognition and Artificial Intelligence · doi:10.1142/s0218001424590110 · Cited by 2
Hofmann, A. B., Dörner, M., Machetanz, L., & Kirchebner, J. (2024) · Healthcare · doi:10.3390/healthcare12171699 · Cited by 2
“This quantification of a predictive model allows a transparent evaluation of the soundness of the model [19].”
MethodBig data and ensemble learning for effective student orientation in Morocco
Badrani, M., Marouan, A., Kannouf, N., & Chetouani, A. (2024) · Indonesian Journal of Electrical Engineering and Computer Science · doi:10.11591/ijeecs.v36.i3.pp1904-1910 · Cited by 3
“Zhou et al. [9] developed a chatbot-based vocational guidance system for Ben M’Sik Faculty of Sciences students.”
BackgroundIgoche, B. I., Matthew, O., Bednár, P., & GEGOV, A. E. (2024) · IEEE International Conference on Intelligent Systems · doi:10.1109/is61756.2024.10705239 · Cited by 3
“By constraining and sparsing the input features through ablation, we aim to reduce the influence of biased features on the LIME explanations while maintaining the explanatory power of the most relevant features [28].”
A Survey on Machine Learning Applications in Renewable Energies Forecasting
Mohabbati, M. (2024) · Power systems · doi:10.1007/978-3-031-69358-8_13 · Cited by 2
Recursive Division Explainability as a Factor of CNN Quality
Gorokhovatskyi, O., & Peredrii, O. (2024) · Lecture notes on data engineering and communications technologies · doi:10.1007/978-3-031-70959-3_16 · Cited by 2
Detection of Thyroid Disease Through Machine Learning Technique
Al‐Momani, M., Al–Qerem, A., Nashwan, S. I., Jebreen, I., Rajab, M. A., Alauthman, M., Jeadi, M., & Elrashidi, A. (2024) · International Arab Conference on Information Technology (ACIT) · doi:10.1109/acit62805.2024.10877258 · Cited by 2
“It offers a broad indicator of a model's performance in every class[35].”
Enhancing Clinical Trust: The Role of AI Explainability in Transforming Healthcare
Pawlicka, A., Pawlicki, M., Jaroszewska-Choraś, D., Kozik, R., & Choraś, M. (2024) · IEEE ... International Conference on Data Mining workshops · doi:10.1109/icdmw65004.2024.00075 · Cited by 3
“…- depending on the source, it is defined as a measure of how accurately a simpler model replicates the behavior of a complex model [29], the percentage of matching predictions between the two models [30], presented as synonymous to the combination of completeness and soundness [31], and so on [32].”
Navigating the informativeness-compression trade-off in XAI
Oldenburg, N., & Søgaard, A. (2025) · AI and Ethics · doi:10.1007/s43681-025-00733-5 · Cited by 4
“3 Overview of the implementation and data processing pipeline queries yield consistent outputs over time [79, 119] (e.g. if a user asks, “Why was this loan denied?” today and then revisits the question next week, the explanation should remain stable unless underlying data or model decisions have…”
Creating Holistic Learning Environments: Integrating AI, Education, and Psychological Principles
Kadir, A., Sairin, S., Madjid, A., Widigdo, M. S. A., Ramli, S. A., & Asrifan, A. (2025) · Advances in computational intelligence and robotics book series · doi:10.4018/979-8-3373-5786-7.ch002 · Cited by 2
Evaluating Explanation Quality in X-IDS Using Feature Alignment Metrics
Alquliti, M., Karafili, E., & Kang, B. (2025) · Lecture notes in computer science · doi:10.1007/978-3-032-00630-1_4 · Cited by 2
Kamath, U., & Liu, J. (2021) · Springer eBooks · doi:10.1007/978-3-030-83356-5_8 · Cited by 2
Rahman, N., Jahan, F., & Alam, F. I. (2022) · International Conference on Innovations in Science, Engineering and Technology (ICISET) · doi:10.1109/iciset54810.2022.9775826 · Cited by 2
A Framework for Auditing Multilevel Models using Explainability Methods
Bhaumik, D., Dey, D., & Kayal, S. (2022) · European Conference on the Impact of Artificial Intelligence and Robotics · doi:10.34190/eciair.4.1.874 · Cited by 2
Scene Text Recognition Models Explainability Using Local Features
Ty, M. V., & Atienza, R. (2023) · Proceedings - International Conference on Image Processing · doi:10.1109/icip49359.2023.10222406 · Cited by 2
“In the current literature, attribution-based explanation methods [9, 16, 15] make up the majority of machine learning explanations.”
BackgroundLohaj, O., Paralič, J., Pella, Z., Pella, D., & Pavlíček, A. (2024) · Diagnostics · doi:10.3390/diagnostics14090917 · Cited by 2
“The new indicators determining the “accuracy” of the system came from the study [18], while “fidelity” and “response time” are mentioned in the other research article [46].”
BackgroundTiny-RFNet: Enabling Modulation Classification of Radio Signals on Edge Systems
Chegini, M., Abdollahi, M., Baniasadi, A., & Patooghy, A. (2024) · CPSSI International Symposium on Cyber-Physical Systems (Applications and Theory) (CPSAT) · doi:10.1109/cpsat64082.2024.10745466 · Cited by 2
“When it comes to measuring the performance of DL models, factors like accuracy, similarity, sensitivity and similar factors can be used [11].”
MethodOn the Objective Evaluation of Post Hoc Explainers.Preprint
Carmichael, Z., & Scheirer, W. J. (2021) · arXiv (Cornell University) · Cited by 4
Characterizing the contribution of dependent features in XAI methodsPreprint
Salih, A. M., Galazzo, I. B., Raisi‐Estabragh, Z., Petersen, S. E., Menegaz, G., & Radeva, P. I. (2023) · arXiv (Cornell University) · doi:10.48550/arxiv.2304.01717 · Cited by 2
Human-Agent Teaming: A System-Theoretic OverviewPreprint
Meimandi, K. J., Bolton, M. L., & Beling, P. A. (2024) · doi:10.36227/techrxiv.170629550.02125312/v1 · Cited by 5
PENERAPAN METODE MACHINE LEARNING UNTUK MENGKLASIFIKASI PENANGANAN PERAWATAN PASIEN
Aziz, F. (2023) · Journal Pharmacy and Application of Computer Sciences · doi:10.59823/jopacs.v1i2.37 · Cited by 1
Case, T. R., Rice, R. C., Cabrera, A., Kricfalusi, M., Dasari, P., Jesurajan, J., Ruckle, D., LaFleur, A., Jobe, C. M., & Syed, H. M. (2023) · International Journal of Research in Orthopaedics · doi:10.18203/issn.2455-4510.intjresorthop20232605 · Cited by 1
A Novel Black-Box Complementary Explanation Approach for Thorax Multi-Label Classification
Bouabdallah, K., Drif, A., & Kaderali, L. (2024) · International Conference on Innovative Research in Applied Science, Engineering and Technology (IRASET) · doi:10.1109/iraset60544.2024.10548809 · Cited by 1
“It is impossible to create quantitative measurements applicable to all interpretability approaches [52].”
BackgroundTowards the Design of Explanation-aware Decision Support Systems
Kolajo, T., & Daramola, O. (2024) · Lecture notes in networks and systems · doi:10.1007/978-3-031-73110-5_7 · Cited by 1
Smart Agriculture Inventory Management System Using Machine Learning Algorithms
Kanyal, H. S., Jain, H., Tyagi, H., & Yadav, S. P. (2024) · International Conference on Computing, Sciences and Communications (ICCSC) · doi:10.1109/iccsc62048.2024.10830398 · Cited by 1
Faithfulness of Local Explanations for Tree-Based Ensemble Models
Rahnama, A. H. A., Geurts, P., & Boström, H. (2025) · Lecture notes in computer science · doi:10.1007/978-3-031-78980-9_2 · Cited by 1
Sharma, S., Kumra, N., Luthra, M., Verma, V., & Sharma, K. (2025) · Wiley eBooks · doi:10.1002/9781394249312.ch19 · Cited by 2
Ogunniyi, J. O., Emuoyibofarhe, J. O., Oladosu, J. B., & Olamoyegun, M. M. (2024) · FUOYE Journal of Engineering and Technology · doi:10.4314/fuoyejet.v9i3.12 · Cited by 1
“Understanding the strengths and limitations of each metric is essential for researchers and healthcare practitioners to make informed decisions when developing and deploying ML-based models (Zhou et al ., 2021).”
BackgroundEvaluation of Explainable Artificial Intelligence for Predictive Process Mining in Education
Oreški, D., & Lukinec, M. P. (2025) · WSEAS TRANSACTIONS ON ADVANCES in ENGINEERING EDUCATION · doi:10.37394/232010.2025.22.1 · Cited by 1
Inam, R., Singh, V., Chawla, A., & Bosneag, A.-M. (2025) · Elsevier eBooks · doi:10.1016/b978-0-44-329135-7.00013-8 · Cited by 1
Data Science Approaches to Analyzing Aesthetic Strategies in Contemporary Presidential Campaigns
Isnawijaya, I. (2025) · Journal of Applied Data Sciences · doi:10.47738/jads.v6i2.609 · Cited by 1
MULTIPLE RECURSIVE DIVISION EXPLANATIONS FOR IMAGE CLASSIFICATION PROBLEMS
Gorokhovatskyi, O., Peredrii, O., & Teslenko, O. (2025) · Advanced Information Systems · doi:10.20998/2522-9052.2025.3.01 · Cited by 2
Evaluation of Local Explainability Methods in Turkish Text Classification Tasks
Porsuk, O., Yıldırım, S., & Başar, A. (2025) · Lecture notes in networks and systems · doi:10.1007/978-3-031-92602-0_30 · Cited by 1
Exploring the Role of Deep Learning in Image Analysis for Breast Cancer Diagnosis
Aitouhanni, I., Bakkali, M. E., Choukri, S., Ouardi, B. E., Mouniane, Y., Masmoudi, S., Rektouti, M., Elorch, I., Belomaria, M., & Aouane, M. (2025) · IGI Global eBooks · doi:10.4018/979-8-3693-8789-4.ch015 · Cited by 1
Ali, U., Nowsherwan, G. A., Riaz, S., & Naseem, S. (2025) · Journal of Computational Electronics · doi:10.1007/s10825-025-02436-w · Cited by 1
“Together, these metrics offer a comprehensive evaluation of model performance, enabling a robust comparison of accuracy, reliability, and (13) generalization across different machine-learning approaches [87].”
Fraud Detection in Financial Transactions Using Ensemble Machine Learning Models
Michael, O., Odeh, C., Maduabuchuku, A., Nwakeze, O. M., Love., E.-O. C., & Akachukwu, O. C. (2025) · International journal of research and scientific innovation · doi:10.51244/ijrsi.2025.1213cs003 · Cited by 1
Sousa, D. P., da Silva, J. M. B., Cavalcante, C. C., & Fischione, C. (2025) · Wiley eBooks · doi:10.1002/9781394249879.ch12 · Cited by 2
WISCA: A Consensus-Based Approach to Harmonizing Interpretability in Tabular Datasets
Banegas‐Luna, A. J., Pérez‐Sánchez, H., & Martínez-Cortés, C. (2026) · Machine Learning and Knowledge Extraction · doi:10.3390/make8040097 · Cited by 1
“However, the utility of XAI is often undercut by the inconsistent explanations these algorithms generate, leading to confusion and diminishing trust Rudin et al. (2022); Zhou et al. (2021); Krishnan (2020); Vowels (2022).”
Zhu, J., Ji, G., Chen, B., Li, X., András, K., Ren, F., Zhu, X., Mu, Z., & Liu, H. (2025) · Journal of Food Composition and Analysis · doi:10.1016/j.jfca.2025.108807 · Cited by 1
ProtoMask: Segmentation-Guided Prototype Learning
Meinert, S., Schlinge, P., Strodthoff, N., & Atzmueller, M. (2025) · IEEE International Conference on Tools with Artificial Intelligence (ICTAI) · doi:10.1109/ictai66417.2025.00128 · Cited by 1
“General studies have demonstrated the complexity and diversity of quantative evaluation techniques employed in our context [35] [36].”
Perspective Chapter: Validation and Quality Metrics of Machine Learning Explanations
García-Cuesta, E., & González, A. M. L. (2026) · IntechOpen eBooks · doi:10.5772/intechopen.1013468 · Cited by 1
NETRA: An AI-Powered Cybercrime Reporting and Analysis System
Aggarwal, C., Singh, V., & Nigam, S. (2025) · IEEE International Conference on Recent Advances in Computing and Systems (REACS) · doi:10.1109/reacs67479.2025.11413391 · Cited by 1
“The primary objective of this study is to evaluate and compare various machine learning and deep learning models for cybercrime classification based on key performance metrics such as accuracy, precision, recall, and F1-score [4], [5].”
DiffSeg: text-guided diffusion-based image editing with semantic segmentation
Li, L., Liu, J., Ye, Z., & Xia, W. (2026) · International Conference on Graphics and Image Processing (ICGIP) · doi:10.1117/12.3095523 · Cited by 1
“• Editing Performance Metrics : – Accuracy [30]: Proportion of correctly predicted samples (both positive and negative) to total samples.”
Analysis of Factors Affecting Defense Expenditures: An XAI-Based Model
KİŞMAN, Z. A., Uzun, S., GÜLER, M. A., Bakal, C., Kıraç, Z. K., Yıldırım, G., Ayden, C., & Alataş, B. (2026) · Systems · doi:10.3390/systems14030283 · Cited by 1
“In practice, the scientific validity of XAI explanations depends on quantitatively testing whether the explanation accurately represents the model and whether feature contribu-https://doi.org/10.3390/systems14030283 tions are truly model-based [40].”
InfluentialBergeron, L., François, J., State, R., & Hilger, J. (2025) · IEEE International Conference on Big Data (BigData) · doi:10.1109/bigdata66926.2025.11402269 · Cited by 1
Yadav, S., Kaushik, A., & McDaid, K. (2026) · Applied AI Letters · doi:10.1002/ail2.70028 · Cited by 1
Enhancing Data Understandability. An Integrated Approach
Mariani, I., Garetto, G., Edmond, J., Lima, E., Santos, M. R. D. A. D. L., Manias, G., Fournier, F., & Limonad, L. (2026) · Springer series in design and innovation · doi:10.1007/978-3-032-23948-8_3 · Cited by 1
Çürük, E., Pınar, M., Pınar, E., & Demirpolat, E. (2026) · Journal of Population Ageing · doi:10.1007/s12062-026-09543-x · Cited by 1
Evaluating Explainability: Metrics, Benchmarks, and Human-Centered Evaluation Methods
Yousefi, N., & Zamanifar, A. (2026) · Springer eBooks · doi:10.1007/978-3-032-15088-2_10 · Cited by 1
Human-Centered Explainable Artificial Intelligence
Zhou, J. J., Chen, F., & Zheng, B. (2026) · Springer eBooks · doi:10.1007/978-981-95-5585-7_11 · Cited by 1
Self-citationIbrahim, A. S. I., Yahaya, A., SAYDI, S., Çomak, G., & Sönmez, Ö. (2026) · Bioresource Technology Reports · doi:10.1016/j.biteb.2026.102832 · Cited by 3
Haas, S., Högemann, M., & Thomas, O. (2026) · Frontiers in Artificial Intelligence · doi:10.3389/frai.2026.1830201 · Cited by 1
“…interviews with larger-scale surveys or experimental studies and complement these qualitative measures with quantitative interaction-based metrics, such as decision accuracy, time required, likelihood to deviate from model recommendations, or the ability to detect model errors (Zhou et al., 2021).”
Explainable lightweight neural networks: A survey on methods, metrics, and challenges
Mushtaq, M., Madden, M. G., & Ullah, I. (2026) · Neurocomputing · doi:10.1016/j.neucom.2026.134504 · Cited by 1
Explainability, Standardisation, and the AI Act
Sapienza, S., & Palmirani, M. (2026) · Lecture notes in computer science · doi:10.1007/978-3-032-34304-8_1 · Cited by 1
Generalization of Machine Learning-Based Compression Method to Hyperspectral Images
Maksimov, A., & Gashnikov, M. V. (2022) · International Conference on Information Technology and Nanotechnology (ITNT) · doi:10.1109/itnt55410.2022.9848734 · Cited by 1
“The main idea is that we can build a current floor predictor: We compute the weight vector 1 F based on the minimum deviation [17-19] between the original HS-floor and the predicted HS-floor:”
BackgroundLopes, B. G. C. O., Soares, L. S., Prates, R. O., & Gonçalves, M. A. (2022) · Journal on Interactive Systems · doi:10.5753/jis.2022.2556 · Cited by 1
“In some works, these concepts are used interchangeably and sometimes even considered synonymous (Mohseni et al., 2021; Vilone and Longo, 2021; Zhou et al., 2021), whereas in others, authors make an explicit distinction between them2.”
BackgroundMetric on the space of systems behavior functions represented by fuzzy measures
Bocharnikov, V., & Sveshnikov, S. (2023) · Yugoslav journal of operations research · doi:10.2298/yjor230315012b · Cited by 1
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.