Shapley Values for Feature Selection: The Good, the Bad, and the Axioms
Open accessThe DOI matches its Crossref record; 63 of the 64 references it cites were found in a registry or an index.
Abstract
The Shapley value has become popular in the Explainable AI (XAI) literature, thanks, to a large extent, to a solid theoretical foundation, including four “favourable and fair” axioms for attribution in transferable utility games. The Shapley value is provably the only solution concept satisfying these axioms. In this paper, we introduce the Shapley value and draw attention to its recent uses as a feature selection tool. We call into question this use of the Shapley value, using simple, abstract “toy” counterexamples to illustrate that the axioms may work against the goals of feature selection. From this, we develop a number of insights that are then investigated in concrete simulation settings, with a variety of Shapley value formulations, including SHapley Additive exPlanations (SHAP) and Shapley Additive Global importancE (SAGE). The aim is not to encourage any use of the Shapley value for feature selection, but we aim to clarify various limitations around their current use in the literature. In so doing, we hope to help demystify certain aspects of the Shapley value axioms that are viewed as “favourable”. In particular, we wish to highlight that the favourability of the axioms depends non-trivially on the way in which the Shapley value is appropriated in the XAI application.
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
“we incorporated the use of interpretability tools (eg, SHAP 28 values, feature importance) to ensure transparency, a priority emphasized in recent ML reporting guidelines.”
From Linguistic markers for identifying post-traumatic stress disorder and associated symptoms: a systematic literature review (2025) · cites this paper as [28]
The check wasn’t confident either way, so it gives no verdict — never disputed.
In the citing paper
“To align with best practices 27 for systematic reviews of ML studies, we augmented PROBAST with criteria tailored to these challenges”
From Linguistic markers for identifying post-traumatic stress disorder and associated symptoms: a systematic literature review (2025) · cites this paper as [28]
Experimental — accuracy not yet measured
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
Gauriat, C.-M., Pencolé, Y., Ribot, P., & Brouillet, G. (2024) · HAL (Le Centre pour la Communication Scientifique Directe) · doi:10.4230/oasics.dx.2024.27 · Cited by 12,586
From local explanations to global understanding with explainable AI for trees
Lundberg, S. M., Erion, G., Chen, H., DeGrave, A., Prutkin, J. M., Nair, B., Katz, R., Himmelfarb, J., Bansal, N., & Lee, S.-I. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-019-0138-9 · Cited by 10,551
17. A Value for n-Person Games
Shapley, L. S. (1953) · Contributions to the Theory of Games (AM-28), Volume II · doi:10.1515/9781400881970-018 · Cited by 6,950
A survey on feature selection methods
Chandrashekar, G., & Sahin, F. (2014) · Computers & Electrical Engineering · doi:10.1016/j.compeleceng.2013.11.024 · Cited by 5,189
Shapley, L. S. (1988) · The Shapley Value · doi:10.1017/cbo9780511528446.003 · Cited by 4,954
Explaining prediction models and individual predictions with feature contributions
Štrumbelj, E., & Kononenko, I. (2014) · Knowledge and Information Systems · doi:10.1007/s10115-013-0679-x · Cited by 2,286
Tukey, J. W. (1962) · The Annals of Mathematical Statistics · doi:10.1214/aoms/1177704711 · Cited by 1,860
Consistent Individualized Feature Attribution for Tree Ensembles
Lundberg, S. M., Erion, G. G., & Lee, S.-I. (2018) · arXiv · doi:10.48550/arxiv.1802.03888 · Cited by 1,430
Budescu, D. V. (1993) · Psychological Bulletin · doi:10.1037/0033-2909.114.3.542 · Cited by 1,193
Research works printed but not linked: 1
- 18. Guyon, U. V. Luxburg, S. Bengio, H. Wallach, R. Fergus, S. Vishwanathan, and R. Garnett, Eds., vol. 30. Curran Associates, Inc., 2017.
Cited by
Items 201–300 of 347
Rubrice, R., Guéneau, V., Briandet, R., Cornuéjols, A., & Guigue, V. (2025) · Artificial Intelligence in the Life Sciences · doi:10.1016/j.ailsci.2025.100137 · Cited by 6
“Indeed, they can be of paramount value but can also be misleading if used or analyzed poorly because of the theoretical assumptions they rely on, such as feature independence [26,45,46].”
Liu, J., Zhou, Z., Chen, Y., Islam, M. T., & Xing, L. (2025) · The American Journal of Human Genetics · doi:10.1016/j.ajhg.2025.08.021 · Cited by 3
Halimi, A., Houmani, N., Garcia-Salicetti, S., Kounis, I., & Coilly, A. (2025) · Health Data Science · doi:10.34133/hds.0295 · Cited by 2
Bromberger, D. A., Júnior, A. N., Garcia, V. J., & Francescatto, M. B. (2025) · Engineering Applications of Artificial Intelligence · doi:10.1016/j.engappai.2025.113558 · Cited by 4
Bae, S., Son, B., Sung, T., Kim, Y., Kim, Y. M., & Im, J. (2025) · GIScience & Remote Sensing · doi:10.1080/15481603.2025.2609352 · Cited by 3
Al-Shqeerat, K. H. A., Abadleh, A., Sharma, S. K., Kumar, P., Tejani, G. G., & Bassir, D. (2026) · Egyptian Informatics Journal · doi:10.1016/j.eij.2025.100874 · Cited by 2
Abreu, M. C., Duarte, R. D. B., Neves, B. D., Silva, N. A., Santos, R., & Campos, J. (2026) · ACM Transactions on Computing for Healthcare · doi:10.1145/3788685 · Cited by 4
“Some approaches transform outputs from traditional XAI techniques, such as SHAP [21] and LIME [46], into more natural and accessible narratives [67].”
A Decision Method for Benefit Distribution Mechanism of Shared End Distribution on Shapley Value
Zhang, H., Hou, Y., & He, F.-F. (2022) · Computational Intelligence and Neuroscience · doi:10.1155/2022/3816440 · Cited by 2
“*e Shapley value is probably the only solution concept satisfying these axioms [43].”
BackgroundRetractedImproving imbalanced dataset classification using Conditional Classifier-Generator (cCGen)
Chakraborty, A., Nadarja, A. V., Milani, A. S., Tobia, J. P., & Narayan, A. (2022) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc53654.2022.9945528 · Cited by 2
“For the Wearable dataset, we used the Shapley values [15] of the features to find those that had the most impact on classification, namely: ”Weight”, ”Age”, ”mean.”
MethodKhamsehashari, R., Schmitt, V., Polzehl, T., Mohtaj, S., & Möller, S. (2023) · Symposium on Security and Privacy in Speech Communication (SPSC) · doi:10.21437/spsc.2023-1 · Cited by 8
“For example, when using COOLANT , MMFN , and TieFake for the fake news detection task and Shapley-based explanations are generated for the text-based outputs [60], they might be not meaningful for users, who do not have any computer science background.”
Feature Selection via Minimal Covering Sets for Industrial Internet of Things Applications
Christou, I. T., Soldatos, J. K., Papadakis, T., Gutiérrez-Rojas, D., & Nardelli, P. H. J. (2023) · International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) · doi:10.1109/dcoss-iot58021.2023.00092 · Cited by 3
“Regarding explaining black-box models, the most common techniques so far have been the SHAP [11] and LIME [12] models that work as follows: SHAP is based on the econometric concept of Shapley values that attempts to award a fair value to the contribution of individual players in a team”
MethodThe Allocations for Dual Fuzzy Cooperative Games
Tang, J., Li, Z., Meng, F., & Tan, C. (2024) · International Journal of Fuzzy Systems · doi:10.1007/s40815-024-01723-1 · Cited by 2
Efficient Shapley performance attribution for least-squares regression
Bell, L., Devanathan, N., & Boyd, S. (2024) · Statistics and Computing · doi:10.1007/s11222-024-10459-9 · Cited by 3
Blaga, A. C., Drăgoi, E.-N., Tucaliuc, A., Kloetzer, L., Puițel, A. C., Caşcaval, D., & Galaction, A. (2024) · Heliyon · doi:10.1016/j.heliyon.2024.e36113 · Cited by 2
“The Min-Max approach [43] was used to normalise the data.”
MethodMaterassi, D., Warnick, S. C., Rojas, C. R., Schoukens, M., & Cross, E. J. (2024) · IFAC-PapersOnLine · doi:10.1016/j.ifacol.2024.08.577 · Cited by 2
A Comprehensive Multifunctional Approach for Measuring Parkinson's Disease Severity
Rahimi, M., Masry, Z. A., Templeton, J. M., Schneider, S. L., & Poellabauer, C. (2024) · Applied Clinical Informatics · doi:10.1055/a-2420-0413 · Cited by 5
Reputation-based Resource Allocation Prioritization in 6G: A Coalitional Game Theoretic Approach
Pathak, V., Sharma, A., Rathore, M. S., & Pandya, R. J. (2024) · International Conference on Computing, Communication and Networking Technologies (ICCCNT) · doi:10.1109/icccnt61001.2024.10726067 · Cited by 2
“…the wireless network, which is given as follows [12] [13]. for an SBS joining a coalition can be calculated based on the shapley value formulation [14]: where ] can be calculated by marginal contribution of the i th BS, before and after joining the coalition, n is total number of SBSs, m is total…”
BackgroundFinancial Data set Classification Method based on Linear Regression Model
Xing, Y. (2024) · Procedia Computer Science · doi:10.1016/j.procs.2024.10.117 · Cited by 2
Drewnick, T., Wen, X., Oberhofer, U., Gorjão, L. R., Beck, C., Hagenmeyer, V., & Schäfer, B. (2025) · Chaos An Interdisciplinary Journal of Nonlinear Science · doi:10.1063/5.0239371 · Cited by 3
Baldoni, S., Benhamadi, S., Chiariotti, F., Zorzi, M., & Battisti, F. (2025) · IEEE Conference on Virtual Reality and 3D User Interfaces (VR) · doi:10.1109/vr59515.2025.00032 · Cited by 2
“To evaluate the influence of the different features on the identification performance, we computed the average Shapley values [9] for the 30-user classification problem, focusing on the best-performing classifiers for each game.”
Bae, S., Son, B., Sung, T., Kang, Y., & Im, J. (2025) · Remote Sensing of Environment · doi:10.1016/j.rse.2025.114735 · Cited by 2
Improved estimation of forage nitrogen in alpine grassland by integrating Sentinel-2 and SIF data
Zhang, Y., Gao, J., Zhang, D., Liang, T., Wang, Z., Zhang, X., Ma, Z., & Yang, J. (2025) · Plant Methods · doi:10.1186/s13007-025-01389-2 · Cited by 2
“” A feature’s SHAP value of a feature represents the average contribution of that feature to all the samples [46].”
An Empirical Evaluation of Explainable AI Approaches
Christou, I. T., Soldatos, J. K., & Lappas, P. Z. (2025) · International Conference on Distributed Computing in Smart Systems and the Internet of Things (DCOSS-IoT) · doi:10.1109/dcoss-iot65416.2025.00159 · Cited by 3
“QARMA ranks first, with the FUZZY approach being a close second; we were a little surprised to see the SHAP approach ranking close, and in fact beating the LIME approach that is closer to a direct interpretation than the “feature contribution” approach of SHAP [20].”
Individualized machine-learning-based clinical assessment recommendation system
Setiawan, D., Wiranto, Y., Girard, J. M., Watts, A. S., & Ashourvan, A. (2025) · PLOS Digital Health · doi:10.1371/journal.pdig.0001022 · Cited by 4
Liu, M., & Zhu, F. (2025) · Artificial Intelligence in Medicine · doi:10.1016/j.artmed.2025.103297 · Cited by 2
Evaluating Interpretability Using Logical Relations: Do Saliency Maps Encode Relevant Information?
Schwenke, L., & Atzmueller, M. (2025) · IEEE International Conference on Data Science and Advanced Analytics (DSAA) · doi:10.1109/dsaa65442.2025.11248004 · Cited by 2
XAI-Assisted Radio Resource Management: Feature Selection and SHAP Enhancement
Ruiz, N. S., Peesapati, S. K. G., & Moysen, J. (2023) · International Conference on Communications · doi:10.1109/icc45041.2023.10279520 · Cited by 3
“The authors in [12] argue that this technique is more suitable in a cooperative game setting and may not be as desirable when applied to ML.”
BackgroundInferring feature importance with uncertainties in high-dimensional dataPreprint
Johnsen, P. V., Strümke, I., Riemer–Sørensen, S., DeWand, A. T., & Langaas, M. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2109.00855 · Cited by 2
“…Lundberg et al. (2020); Sellereite and Jullum (2019); Lundberg and Lee (2017); Strumbelj and Kononenko (2013, 2010); Lundberg et al. (2019); Redelmeier et al. (2020); Kwon et al. (2021); Song et al. (2016); Moehle et al. (2021); Covert et al. (2020a); Keinan et al. (2003); Fryer et al. (2021b)).”
BackgroundMerchant, M. A., McBlane, L., & Edwards, R. (2024) · IEEE Mediterranean and Middle-East Geoscience and Remote Sensing Symposium (M2GARSS) · doi:10.1109/m2garss57310.2024.10537286 · Cited by 1
Cybersecurity Analysis Through Shapley Values for a Network Traffic Dataset of Android Malware
Nikolaeva, D., Buratto, A., & Badia, L. (2024) · International Conference on Communications and Signal Processing (ICCSP) · doi:10.1109/iccsp60870.2024.10543711 · Cited by 2
“The core of the analysis is the application of Shapley values [10], [11], derived from cooperative game theory and quantified through the SHapley Additive exPlanations (SHAP) toolbox [12], to evaluate the contribution of individual data points.”
MethodMachine Learning within Latent Spaces formed by Foundation Models
Tomczyk, B., Angelov, P. P., & Kangin, D. (2024) · IEEE International Conference on Intelligent Systems · doi:10.1109/is61756.2024.10705264 · Cited by 2
A Novel Choquet Integral-Based Regression Approach for Analyzing Freight-Train Derailment Severity
Huang, G., Li, H., Gao, X., Guo, Z., Wang, G., Zhao, H., & Song, Z. (2025) · Transportation Research Record Journal of the Transportation Research Board · doi:10.1177/03611981251318327 · Cited by 1
“Within the realm of machine learning, this method has been used for feature selection ( 49 , 50 ) and pruning ( 51 , 52 ).”
Yu, G., Zhuo, Q., & Wang, C. (2025) · Journal of Environmental Science and Health Part B · doi:10.1080/03601234.2025.2489259 · Cited by 1
“In 2017, Lundberg et al., [42] drawing upon L.S. Shapley’s game theory concept of “fair distribution”, [43] introduced the SHAP method for model interpretation.”
Fan, S. S., & Zheng, A.-T. (2025) · IEEE Transactions on Automation Science and Engineering · doi:10.1109/tase.2025.3570863 · Cited by 3
“Recursive feature elimination (RFE) based on the SH apley A dditive ex P lanations (SHAP) value is applied to numerical features [25], [26].”
Evaluating Aesthetics of Arabic Handwriting: Hand Motion-Based Approach
Babushkin, V., Alsuradi, H., Al-Khalil, M. O., & Eid, M. A. (2025) · Lecture notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering · doi:10.1007/978-3-031-97254-6_7 · Cited by 1
Fast Approximation of Shapley Values Through Fractional Factorial Designs
Zhou, Z., Mee, R. W., Hamers, H., & Zheng, W. (2025) · Journal of the American Statistical Association · doi:10.1080/01621459.2025.2529027 · Cited by 1
“For example, feature selection tasks require building a new model and making predictions for the entire dataset (Fryer, Strümke, and Nguyen 2021).”
Ng, K. W., Huang, Y. F., Koo, C. H., El‐Shafie, A., & Ahmed, A. N. (2025) · Hydrological Sciences Journal · doi:10.1080/02626667.2025.2539843 · Cited by 1
Fosic, I., Starčević, D. P., & Fosic, I. (2025) · International Convention on Information and Communication Technology, Electronics and Microelectronics · doi:10.1109/mipro65660.2025.11131973 · Cited by 1
Pseudo datasets estimate feature attribution in artificial neural networks
Yang, H.-Y., Chen, Y., Cheng, H., & Guo, C. (2025) · Scientific Reports · doi:10.1038/s41598-025-28840-7 · Cited by 1
“In a further literature review, Fryer et al. (2021) discussed the application of Shapley value in feature selection in detail 37 .”
Blaga, A. C., Drăgoi, E.-N., Caşcaval, D., Puițel, A. C., & Galaction, A. (2025) · Scientific Reports · doi:10.1038/s41598-025-26825-0 · Cited by 1
“Figure 12 presents the SHAP values, where the positive values on the x-axis indicate that that specific input leads to a rising trend of the output.”
InfluentialExpulsion Prediction in Resistance Spot Welding – Process Optimization with Explainability Methods
Durnagöz, S., Portnoy, S. M., Mayer, M., & Huber, M. F. (2026) · Procedia CIRP · doi:10.1016/j.procir.2026.01.127 · Cited by 1
Feature selection with a lexicographic social ranking method
Gourvès, L., Moretti, S., & Tamby, S. (2026) · International Journal of Approximate Reasoning · doi:10.1016/j.ijar.2026.109656 · Cited by 3
The LASSO path as a feature importance proxy
Fernandez-Piana, L., & Moreno, V. (2026) · International Journal of Data Science and Analytics · doi:10.1007/s41060-026-01094-w · Cited by 1
Fan, Y., Tian, J., Sun, D., Li, S., Liu, M., Ji, Q., & Yin, W. (2026) · npj Heritage Science · doi:10.1038/s40494-026-02473-z · Cited by 1
Wu, H., Weng, J., & Zeng, Y. (2026) · Biomedical Signal Processing and Control · doi:10.1016/j.bspc.2026.110382 · Cited by 1
A Comparison of Feature Selection Using SHAP-Values for Tree and Non-tree-Based Models
Santavicca, F., Rietdijk, H. H., Conde-Céspedes, P., & Trocan, M. (2026) · Communications in computer and information science · doi:10.1007/978-981-92-0068-9_22 · Cited by 1
Peer-to-Peer Federated Learning: A Comprehensive Survey
Allen, A., Mylonas, A., Vidalis, S., & Pitropakis, N. (2026) · AI · doi:10.3390/ai7070243 · Cited by 2
“Readers interested in the Shapley value can refer to [50].”
Wu, J., Liu, S., Huang, X., Zhang, Y., Qin, X., Zhu, W., & Cheng, R. (2026) · Forests · doi:10.3390/f17080874 · Cited by 1
Koç, K. (2023) · Journal of Construction Engineering Management & Innovation · doi:10.31462/jcemi.2023.01001015 · Cited by 2
EvidenceQuest: An Interactive Evidence Discovery System for Explainable Artificial Intelligence
Hanif, A., Beheshti, A., Zhang, X., Wood, S., Benatallah, B., & Foo, J. (2024) · ACM International Conference on Web Search and Data Mining (WSDM) · doi:10.1145/3616855.3635697 · Cited by 1
“Previous studies have used feature selection techniques to improve ML models [6, 9], and our proposed approach investigates the evolution of feature contribution, which can aid in selecting the most relevant features for the model using existing techniques [17, 19].”
MethodData Exchange Markets via Utility Balancing
Bhaskara, A., Gollapudi, S., Im, S., Kollias, K., Munagala, K., & Sankar, G. S. (2024) · International World Wide Web Conference (WWW) · doi:10.1145/3589334.3645364 · Cited by 3
“Though this method has its roots in cost sharing in Economics, it has seen a resurgence in interest as a method to measure the utility of individual datasets for a machine learning task [17, 18].”
BackgroundMethodInfluentialEarly prediction of milk yield from Holstein-Friesian cow lactation in Querétaro, Mexico
Guevara-Escobar, A., Cervantes‐Jiménez, M., García-Muñiz, J. G., Lemus-Ramírez, V., & Yabuta-Osorio, A. K. (2024) · Agro Productividad · doi:10.32854/agrop.v17i7.2743 · Cited by 1
Unlocking the Potential of Patient Metadata for Skin Cancer Detection: An AI Framework
Islam, M. S., Wishart, G. C., Walls, J., Hall, P. N., Garcia, A., Gan, J. Q., & Raza, H. (2024) · IEEE International Conference on Digital Health (ICDH) · doi:10.1109/icdh62654.2024.00026 · Cited by 1
“One such method frequently used for the interpretability of ML models is known as Shapely Additive Explanations (SHAP) [15].”
MethodLeveraging Handwriting Dynamics, Explainable AI and Machine Learning for Alzheimer Prediction
Singh, S. K., & Chaturvedi, A. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-81342-9_27 · Cited by 3
Explainable Predictive Factors for Inter-Agency Partnership Success
Drago, N., Braz, G. A., & Vardanega, T. (2025) · IEEE Access · doi:10.1109/access.2025.3547439 · Cited by 1
Improved Sampling Methods for Evaluation of Classification Performance
Lee, Y., Yen, S.-J., & Tang, Y.-J. (2025) · International Conference on Artificial Intelligence in Information and Communication (ICAIIC) · doi:10.1109/icaiic64266.2025.10920770 · Cited by 1
“Shapely value is a concept in cooperative game theory that can be applied to machine learning to explain the predictions of the model [17].”
Morais, L. R. D. A., Fragoso, G. A. M., Ludermir, T. B., & Monteiro, C. L. A. (2024) · doi:10.5753/eniac.2024.244444 · Cited by 1
Explainable Multi-label Classification for Predictive Maintenance
Hubermont, A., Vellinger, A., & Tuci, E. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-93631-9_15 · Cited by 1
Zhong, S., Pan, Y., & Bian, L. (2025) · International Journal of Geographical Information Systems · doi:10.1080/13658816.2025.2564772 · Cited by 1
“2019, Fryer et al .”
Analysis of Lifelong for Health of Middle-Aged Men by Using Machine Learning Algorithm
Kim, J., Lee, J., & Park, M. (2021) · Journal of Korean Institute of Industrial Engineers · doi:10.7232/jkiie.2021.47.6.504 · Cited by 1
Shapley Value in Partition Function Form Games: New Research Perspectives for Features Selection
Bimonte, G., & Senatore, L. (2022) · Springer eBooks · doi:10.1007/978-3-030-99638-3_17 · Cited by 1
Amoukou, S. I., Brunel, N., & Salaün, T. (2021) · arXiv (Cornell University) · Cited by 1
Zola, F. (2022) · Academica-e (Universidad Pública de Navarra) · doi:10.48035/tesis/2454/44259 · Cited by 1
Chatterjee, A., Ganesh, K., Reigler, M., & Halvorsen, P. (2023) · Research Square · doi:10.21203/rs.3.rs-3114142/v3 · Cited by 1
Predicting Postoperative Delirium in Older Patients: a multicenter retrospective cohort studyPreprint
Wu, S.-C. J., Sharma, N., Bauch, A., Yang, H.-C., Hect, J. L., Thomas, C., Wagner, S., Förstner, B. R., Arnim, C. A. F. V., Kaufmann, T., et al. (2024) · medRxiv · doi:10.1101/2024.03.13.24303920 · Cited by 1
Wegner, P., Grobe‐Einsler, M., Reimer, L. M., Kahl, F., Koyak, B., Elter, T., Lange, A., Kimmich, O., Soub, D., Hufschmidt, F., et al. (2024) · medRxiv · doi:10.1101/2024.05.29.24308057 · Cited by 1
“Incorporating SHAP for model explainability has become a popular choice in machine learning (Fryer et al., 2021) and hence was also used in this work.”
MethodRecent advances in formal explainability
Huang, X. (2023) · theses.fr (ABES) · doi:10.70675/1b6e4710z8c13z4f84zbb0az80e114bebb07 · Cited by 2
(2021) · PeerJ Computer Science · doi:10.7287/peerj-cs.479v0.2/reviews/1
“However, since the values of φ already represent a ranking of the features importance, the user could keep the K -topmost features, and drop the others to obtain a concise explanation (Fryer et al., 2021).”
(2022) · Geoscientific model development · doi:10.5194/gmd-2022-134-rc1
Jin, J. (2022) · Geoscientific model development · doi:10.5194/gmd-2022-134-ac2
Jin, J. (2022) · Geoscientific model development · doi:10.5194/gmd-2022-134-ac1
(2022) · Geoscientific model development · doi:10.5194/gmd-2022-134-rc2
Andreoletti, D., Ayoub, O., Peternier, A., Leidi, T., & Giordano, S. (2022) · IEEE Intl Conf on Dependable, Autonomic and Secure Computing, Intl Conf on Pervasive Intelligence and Computing, Intl Conf on Cloud and Big Data Computing, Intl Conf on Cyber Science and Technology Congress (DASC/PiCom/CBDCom/CyberSciTech) · doi:10.1109/dasc/picom/cbdcom/cy55231.2022.9927830 · Cited by 1
“In the context of feature selection, Shapley values are proving to be an effective approach in accurately estimating features’ contributions and hence, features’ importance [16]–[18].”
BackgroundChatterjee, A., Ganesh, K., Reigler, M., & Halvorsen, P. (2023) · Research Square · doi:10.21203/rs.3.rs-3114142/v1
de Hauteclocque, A. B., Ferrer, L., Margue, G., Etchepare, G., Colin, T., & Bernhard, J. (2023) · British Journal of Urology · doi:10.1111/bju.16116
Li, M., Xie, R., & Tian, J. (2023) · International Conference on Applied Mathematics, Modelling and Intelligent Computing (CAMMIC) · doi:10.1117/12.2685933
“2 Introduction of Shapley value The Shapley value is a common evaluation metric because it provides a fair approach to score features [11].”
BackgroundChatterjee, A., Ganesh, K., Reigler, M., & Halvorsen, P. (2023) · Research Square · doi:10.21203/rs.3.rs-3114142/v2 · Cited by 1
Muñoz-Organero, M. (2023) · Mathematics · doi:10.3390/math11183904
Improving Explainable AI Interpretability: Mathematical Models for Evaluating Explanation Methods.Preprint
Nancy Ambritta, P., Mahalle, P. N., Bhapkar, H. R., Shinde, G. R., & Sable, N. P. (2023) · Research Square · doi:10.21203/rs.3.rs-3528413/v1
Colorado, V., Dubois, P. A., Juma, M., Topsakal, O., & Akinci, T. C. (2023) · International Conference on Engineering and Applied Natural Sciences (ICEANS) · doi:10.59287/as-proceedings.275
Kumar, E. A. S. S. (2023) · International Journal on Recent and Innovation Trends in Computing and Communication · doi:10.17762/ijritcc.v11i10.8481
Souza, L., de Azevedo, K. S., Dalmolin, M., & Fernandes, M. A. C. (2023) · Congresso Brasileiro de Inteligência Computacional (CBIC) · doi:10.21528/cbic2023-093
First Lactation Milk Yield Predicted by the Heifer’s Growth Curve Derivatives
Guevara-Escobar, A., Cervantes‐Jiménez, M., Lemus-Ramírez, V., García-Muñiz, J. G., & Yabuta-Osorio, A. K. (2024) · Dairy · doi:10.3390/dairy5020020
“However, there is currently a growing focus on the models generated The SHapley Additive exPlanations (SHAP) [32] values are an option for model interpretability; the SHAP values assist in post hoc agnostic interpretability of the variable importance in models. by machine learning [31].”
BackgroundDecomposing the Life Expectancy premiumPreprint
Boccard, N. (2024) · SSRN Electronic Journal · doi:10.2139/ssrn.4840447
Evaluation of Classical and Deep Learning Approaches for Human Activity Recognition
Lončarević, Z., Luštrek, M., & Gams, A. (2024) · International Convention on Information and Communication Technology, Electronics and Microelectronics · doi:10.1109/mipro60963.2024.10569489
“In the future, we plan to evaluate and compare these methods against a newer method called Shapley values [22] that was shown to be useful in certain examples [23].”
MethodIdentification of Users in a Gambling Problem with the Use of Machine Learning
Jach, T., Probierz, B., Kozák, J. T., Stefański, P., Dziczkowski, G., Hrabia, A., Juszczuk, P., Głowania, S., Wołek, G., Sznapka, W., et al. (2024) · Lecture notes in computer science · doi:10.1007/978-981-97-4985-0_21 · Cited by 1
Games, Fuzzy Measures, Indices, and Explainable ML: Exploiting the Game
Torra, V. (2024) · Lecture notes in networks and systems · doi:10.1007/978-3-031-70018-7_4
Wu, S.-C. J., Sharma, N., Bauch, A., Yang, H.-C., Hect, J. L., Thomas, C., Wagner, S., Förstner, B. R., Arnim, C. A. F. V., Kaufmann, T., et al. (2024) · doi:10.2196/preprints.67958
Arantes, R. F. M. (2024) · Digital Library of Theses and Dissertations (Universidade de São Paulo) · doi:10.11606/t.18.2024.tde-13012025-162657 · Cited by 2
Shapley-Based Feature Selection for Online Algorithm Selection
Becker, P., & Bengs, V. (2025) · Communications in computer and information science · doi:10.1007/978-3-031-74630-7_21
“A priori” shapley data value estimation
Navia-Vázquez, A., Cid‐Sueiro, J., & Vázquez, M. A. (2025) · Pattern Analysis and Applications · doi:10.1007/s10044-025-01454-5 · Cited by 3
“The Shapley approach offers a well-known solution with multiple applications in ML such as feature selection, model interpretation, robustness against malicious participants or data valuation itself [17, 47, 51, 59].”
Meteorological Factors and the Spread of COVID‐19: A Territorial Analysis in Italy
Telesca, V., Castronuovo, G., Gianfranco, F., Mariarosaria, M., Rondinone, M., & Ceppi, A. (2025) · Meteorological Applications · doi:10.1002/met.70048
Machine Learning-based Electrical Energy Consumption Prediction
Yoon, S., Oh, S., Kim, J., & Park, M. (2025) · Journal of Korean Institute of Industrial Engineers · doi:10.7232/jkiie.2025.51.2.171
Evaluating Shap Xai for Feature Selection in Deep Learning Based Grid-Scale Demand ForecastingPreprint
van Zyl, C., Ye, X., Naidoo, R. M., & Chen, Y. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5244861
Kashyap, S., Shukla, A. K., & Naim, I. (2025) · Elsevier eBooks · doi:10.1016/b978-0-443-23979-3.00009-9
Metadata
- Authors
- Fryer, D., Strumke, I., & Nguyen, H.
- Affiliations
- The University of Queensland (AU)
- La Trobe University (AU)
- Venue
- IEEE Access
- Published
- 2021
- Volume
- 9
- Pages
- 144352-144360
- ISSN
- 2169-3536 (ISSN-L)
- OpenAlex
- W3131037858
- Semantic Scholar
- 6573fa74af238d1e3538c026997e31b9f67f19f7
- Cited by
- 347
- References
- 64
- Citations by year
- 2026: 56 · 2025: 102 · 2024: 86 · 2023: 58 · 2022: 34 · 2021: 11
- FWCI
- 16.63
- Type
- journal article
- Language
- en
- Open access
- gold
- doi.org
- Licence
- Locations
- IEEE Access (publishedVersion, cc-by)
- arXiv (Cornell University) (submittedVersion)
- DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
- Figshare (submittedVersion, cc-by)
- arXiv (Cornell University)
- La Trobe University (cc-by)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Explainable Artificial Intelligence (XAI)
- Social Sciences › Economics, Econometrics and Finance › Economics and Econometrics › Game Theory and Voting Systems
- Physical Sciences › Computer Science › Artificial Intelligence › Privacy-Preserving Technologies in Data
- Keywords
- Shapley value, feature selection, explainable AI, axioms, SHAP, sage, counterexample, transferable utility games, attribution
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.