{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T21:54:09Z","timestamp":1781646849066,"version":"3.54.5"},"reference-count":40,"publisher":"Springer Science and Business Media LLC","issue":"1","license":[{"start":{"date-parts":[[2026,3,1]],"date-time":"2026-03-01T00:00:00Z","timestamp":1772323200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T00:00:00Z","timestamp":1774828800000},"content-version":"vor","delay-in-days":29,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0"}],"funder":[{"name":"Karlsruher Institut f\u00fcr Technologie (KIT)"}],"content-domain":{"domain":["link.springer.com"],"crossmark-restriction":false},"short-container-title":["K\u00fcnstl Intell"],"published-print":{"date-parts":[[2026,3]]},"abstract":"<jats:title>Abstract<\/jats:title>\n                  <jats:p>\n                    Machine learning models are increasingly deployed in critical energy infrastructure, where domain experts require transparent explanations for decision-making. SHAP (SHapley Additive exPlanations) has become a popular method for energy systems applications. However, energy data exhibit inherent correlations due to physical constraints, operational relationships, and market dynamics, posing challenges for interpreting SHAP-based explanations. This work investigates how feature correlations influence SHAP-based explanations using controlled synthetic experiments and real-world power grid data. Our analysis shows that only correlation-aware methods can attribute importance to economically linked features, such as solar generation in predicting fossil fuels, which may reflect genuine systemic interdependencies that are valuable for prediction and scientific understanding. Our findings highlight the tradeoff between\n                    <jats:italic>true to the model<\/jats:italic>\n                    explanations that reflect model behavior and\n                    <jats:italic>true to the data<\/jats:italic>\n                    approaches that consider real-world dependencies. In complex energy systems with circular dependencies, temporal dynamics, and hidden constraints, explanation validity cannot be universally defined. We emphasize the need for practitioners\u2019 awareness of the trade-offs between model analysis, scientific discovery, and operational understanding.\n                  <\/jats:p>","DOI":"10.1007\/s13218-026-00904-4","type":"journal-article","created":{"date-parts":[[2026,3,30]],"date-time":"2026-03-30T12:21:03Z","timestamp":1774873263000},"page":"41-54","update-policy":"https:\/\/doi.org\/10.1007\/springer_crossmark_policy","source":"Crossref","is-referenced-by-count":0,"title":["Correlations Matter in Explanations for Energy Systems"],"prefix":"10.1007","volume":"40","author":[{"given":"Alexandra","family":"Nikoltchovska","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Sebastian","family":"P\u00fctz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Markus","family":"G\u00f6tz","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Benjamin","family":"Sch\u00e4fer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"297","published-online":{"date-parts":[[2026,3,30]]},"reference":[{"key":"904_CR1","doi-asserted-by":"publisher","unstructured":"Aas K, Jullum M, L\u00f8land A (2020) Explaining individual predictions when features are dependent: more accurate approximations to Shapley values. https:\/\/doi.org\/10.48550\/arXiv.1903.10464. ArXiv:1903.10464 [stat]","DOI":"10.48550\/arXiv.1903.10464"},{"key":"904_CR2","doi-asserted-by":"publisher","unstructured":"Antonopoulos I, Robu V, Couraud B, Flynn D (2021) Data-driven modelling of energy demand response behaviour based on a large-scale residential trial. Energy AI 4:100071. https:\/\/doi.org\/10.1016\/j.egyai.2021.100071https:\/\/www.sciencedirect.com\/science\/article\/pii\/S2666546821000252","DOI":"10.1016\/j.egyai.2021.100071"},{"issue":"1","key":"904_CR3","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1186\/s42162-025-00524-6","volume":"8","author":"S Aslam","year":"2025","unstructured":"Aslam S, Aung PP, Rafsanjani AS, Majeed APPA (2025) Machine learning applications in energy systems: current trends, challenges, and research directions. Energy Inform 8(1):62. https:\/\/doi.org\/10.1186\/s42162-025-00524-6","journal-title":"Energy Inform"},{"key":"904_CR4","unstructured":"Berlin DIW DIW Berlin: Residual load, renewable surplus generation and storage requirements in Germany. https:\/\/www.diw.de\/de\/diw_01.c.458121.de\/publikationen\/diskussionspapiere\/2013_1316\/residual_load__renewable_surplus_generation_and_storage_requirements_in_germany.html"},{"issue":"6","key":"904_CR5","doi-asserted-by":"publisher","first-page":"590","DOI":"10.1038\/s42256-023-00657-x","volume":"5","author":"H Chen","year":"2023","unstructured":"Chen H, Covert IC, Lundberg SM, Lee SI (2023) Algorithms to estimate Shapley value feature attributions. Nat Mach Intell 5(6):590\u2013601. https:\/\/doi.org\/10.1038\/s42256-023-00657-x (https:\/\/www.nature.com\/articles\/s42256-023-00657-x)","journal-title":"Nat Mach Intell"},{"key":"904_CR6","doi-asserted-by":"publisher","unstructured":"Chen H, Janizek JD, Lundberg S, Lee SI (2020) True to the model or true to the data?. https:\/\/doi.org\/10.48550\/arXiv.2006.16234. ArXiv:2006.16234 [cs, stat]","DOI":"10.48550\/arXiv.2006.16234"},{"key":"904_CR7","doi-asserted-by":"publisher","unstructured":"Chen T, Guestrin C (2016) XGBoost: a scalable tree boosting system. In: Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pp 785\u2013794. https:\/\/doi.org\/10.1145\/2939672.2939785. ArXiv:1603.02754 [cs]","DOI":"10.1145\/2939672.2939785"},{"key":"904_CR8","unstructured":"ENTSO-E (2025) Entso-e transparency platform. https:\/\/transparency.entsoe.eu\/. Accessed 12 Aug 2025"},{"key":"904_CR9","unstructured":"ENTSO-E (2025) Transparency platform. https:\/\/transparency.entsoe.eu"},{"key":"904_CR10","doi-asserted-by":"publisher","first-page":"144352","DOI":"10.1109\/ACCESS.2021.3119110","volume":"9","author":"D Fryer","year":"2021","unstructured":"Fryer D, Str\u00fcmke I, Nguyen H (2021) Shapley values for feature selection: the good, the bad, and the axioms. IEEE Access 9:144352\u2013144360","journal-title":"IEEE Access"},{"key":"904_CR11","unstructured":"Gerber E (2021) Radical Shapley values. https:\/\/edden-gerber.github.io\/shapley-part-2\/. GitHub: https:\/\/github.com\/edden-gerber\/radical-shapley-values"},{"key":"904_CR12","doi-asserted-by":"publisher","unstructured":"Heskes T, Sijben E, Bucur IG, Claassen T (2020) Causal Shapley values: exploiting causal knowledge to explain individual predictions of complex models. https:\/\/doi.org\/10.48550\/arXiv.2011.01625. ArXiv:2011.01625 [cs]","DOI":"10.48550\/arXiv.2011.01625"},{"key":"904_CR13","doi-asserted-by":"publisher","first-page":"1054","DOI":"10.1016\/j.apenergy.2018.04.048","volume":"225","author":"L Hirth","year":"2018","unstructured":"Hirth L, M\u00fchlenpfordt J, Bulkeley M (2018) The ENTSO-E transparency platform\u2014a review of Europe\u2019s most ambitious electricity data platform. Appl Energy 225:1054\u20131067. https:\/\/doi.org\/10.1016\/j.apenergy.2018.04.048 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S0306261918306068)","journal-title":"Appl Energy"},{"key":"904_CR14","doi-asserted-by":"publisher","first-page":"13960","DOI":"10.1109\/ACCESS.2019.2894819","volume":"7","author":"E Hossain","year":"2019","unstructured":"Hossain E, Khan I, Un-Noor F, Sikander SS, Sunny MSH (2019) Application of big data and machine learning in smart grid, and associated security concerns: a review. IEEE Access 7:13960\u201313988. https:\/\/doi.org\/10.1109\/ACCESS.2019.2894819 (https:\/\/ieeexplore.ieee.org\/document\/8625421)","journal-title":"IEEE Access"},{"key":"904_CR15","doi-asserted-by":"publisher","first-page":"112000","DOI":"10.1016\/j.rser.2021.112000","volume":"156","author":"OT Huxley","year":"2022","unstructured":"Huxley OT, Taylor J, Everard A, Briggs J, Tilley K, Harwood J, Buckley A (2022) The uncertainties involved in measuring national solar photovoltaic electricity generation. Renew Sustain Energy Rev 156:112000. https:\/\/doi.org\/10.1016\/j.rser.2021.112000 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S1364032121012636)","journal-title":"Renew Sustain Energy Rev"},{"key":"904_CR16","unstructured":"(IEA) IEA (2024) World energy outlook 2024\u2014analysis. https:\/\/www.iea.org\/reports\/world-energy-outlook-2024"},{"key":"904_CR17","doi-asserted-by":"publisher","unstructured":"Janzing D, Minorics L, Bl\u00f6baum P (2019) Feature relevance quantification in explainable AI: a causal problem. https:\/\/doi.org\/10.48550\/arXiv.1910.13413. ArXiv:1910.13413 [stat]","DOI":"10.48550\/arXiv.1910.13413"},{"key":"904_CR18","doi-asserted-by":"publisher","unstructured":"Kaplan J, McCandlish S, Henighan T, Brown TB, Chess B, Child R, Gray S, Radford A, Wu J, Amodei D (2020) Scaling laws for neural language models. https:\/\/doi.org\/10.48550\/arXiv.2001.08361. ArXiv:2001.08361 [cs]","DOI":"10.48550\/arXiv.2001.08361"},{"issue":"11","key":"904_CR19","doi-asserted-by":"publisher","first-page":"100365","DOI":"10.1016\/j.patter.2021.100365","volume":"2","author":"J Kruse","year":"2021","unstructured":"Kruse J, Sch\u00e4fer B, Witthaut D (2021) Revealing drivers and risks for power grid frequency stability with explainable AI. Patterns 2(11):100365. https:\/\/doi.org\/10.1016\/j.patter.2021.100365 (https:\/\/linkinghub.elsevier.com\/retrieve\/pii\/S2666389921002270)","journal-title":"Patterns"},{"key":"904_CR20","doi-asserted-by":"publisher","first-page":"108603","DOI":"10.1016\/j.est.2023.108603","volume":"72","author":"P Liu","year":"2023","unstructured":"Liu P, Trieb F (2023) German atlas of thermal storage power plants (TSPP) - a first approach. J Ener Storage 72:108603. https:\/\/doi.org\/10.1016\/j.est.2023.108603 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S2352152X23020005)","journal-title":"J Ener Storage"},{"key":"904_CR21","doi-asserted-by":"publisher","unstructured":"Lundberg SM, Erion G, Chen H, DeGrave A, Prutkin JM, Nair B, Katz R, Himmelfarb J, Bansal N, Lee SI (2019) Explainable AI for trees: from local explanations to global understanding. https:\/\/doi.org\/10.48550\/arXiv.1905.04610. ArXiv:1905.04610 [cs, stat]","DOI":"10.48550\/arXiv.1905.04610"},{"key":"904_CR22","doi-asserted-by":"publisher","unstructured":"Lundberg SM, Lee SI (2017) A unified approach to interpreting model predictions. In: Proceedings of the 31st international conference on neural information processing systems, NIPS\u201917. Curran Associates Inc., Red Hook, NY, USA, pp 4768\u20134777. https:\/\/doi.org\/10.48550\/arXiv.1705.07874","DOI":"10.48550\/arXiv.1705.07874"},{"key":"904_CR23","unstructured":"Molnar C (2023) Interpreting machine learning models with SHAP. A guide with python examples and theory on Shapley values, 1 edn. https:\/\/christophmolnar.com\/books\/shap"},{"key":"904_CR24","unstructured":"Molnar C (2025) Interpretable machine learning. A guide for making black box models explainable, 3 edn. https:\/\/christophm.github.io\/interpretable-ml-book"},{"key":"904_CR25","doi-asserted-by":"publisher","unstructured":"Monti A, Ponci F (2015) Electric power systems. In: Kyriakides E, Polycarpou M (eds) Intelligent monitoring, control, and security of critical infrastructure systems. Springer, Berlin, Heidelberg, pp 31\u201365. https:\/\/doi.org\/10.1007\/978-3-662-44160-2_2","DOI":"10.1007\/978-3-662-44160-2_2"},{"issue":"1","key":"904_CR26","doi-asserted-by":"publisher","first-page":"e0315955","DOI":"10.1371\/journal.pone.0315955","volume":"20","author":"HN Nguyen","year":"2025","unstructured":"Nguyen HN, Tran QT, Ngo CT, Nguyen DD, Tran VQ (2025) Solar energy prediction through machine learning models: a comparative analysis of regressor algorithms. PLoS ONE 20(1):e0315955. https:\/\/doi.org\/10.1371\/journal.pone.0315955","journal-title":"PLoS ONE"},{"key":"904_CR27","unstructured":"Nikoltchovska A, P\u00fctz S, G\u00f6tz M, Sch\u00e4fer B (2025) Correlations-xai-energy. https:\/\/github.com\/KIT-IAI-DRACOS\/correlations-xai-energy"},{"key":"904_CR28","doi-asserted-by":"publisher","unstructured":"Nikoltchovska A, P\u00fctz S, Li X, Hagenmeyer V, Sch\u00e4fer B (2025) Probabilistic and explainable machine learning for tabular power grid data. In: Proceedings of the 16th ACM international conference on future and sustainable energy systems, e-energy \u201925. Association for Computing Machinery, New York, NY, USA, pp 213\u2013231. https:\/\/doi.org\/10.1145\/3679240.3734623","DOI":"10.1145\/3679240.3734623"},{"issue":"4","key":"904_CR29","doi-asserted-by":"publisher","first-page":"1782","DOI":"10.1007\/s10618-024-01016-z","volume":"38","author":"LHB Olsen","year":"2024","unstructured":"Olsen LHB, Glad IK, Jullum M, Aas K (2024) A comparative study of methods for estimating model-agnostic Shapley value explanations. Data Min Knowl Disc 38(4):1782\u20131829. https:\/\/doi.org\/10.1007\/s10618-024-01016-z","journal-title":"Data Min Knowl Disc"},{"key":"904_CR30","doi-asserted-by":"publisher","unstructured":"Owen G (1977) Values of games with a priori unions. In: Henn R, Moeschlin O (eds) Mathematical economics and game theory. Springer, Berlin, Heidelberg, pp 76\u201388. https:\/\/doi.org\/10.1007\/978-3-642-45494-3_7","DOI":"10.1007\/978-3-642-45494-3_7"},{"issue":"7642","key":"904_CR31","doi-asserted-by":"publisher","first-page":"393","DOI":"10.1038\/542393a","volume":"542","author":"S Pfenninger","year":"2017","unstructured":"Pfenninger S (2017) Energy scientists must show their workings. Nature 542(7642):393\u2013393. https:\/\/doi.org\/10.1038\/542393a (https:\/\/www.nature.com\/articles\/542393a)","journal-title":"Nature"},{"key":"904_CR32","doi-asserted-by":"publisher","first-page":"211","DOI":"10.1016\/j.enpol.2016.11.046","volume":"101","author":"S Pfenninger","year":"2017","unstructured":"Pfenninger S, DeCarolis J, Hirth L, Quoilin S, Staffell I (2017) The importance of open data and software: is energy research lagging behind? Energy Policy 101:211\u2013215. https:\/\/doi.org\/10.1016\/j.enpol.2016.11.046 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S0301421516306516)","journal-title":"Energy Policy"},{"key":"904_CR33","doi-asserted-by":"publisher","unstructured":"P\u00fctz S, Kruse J, Witthaut D, Hagenmeyer V, Sch\u00e4fer B (2023) Regulatory changes in German and Austrian power systems explored with explainable artificial intelligence. In: Companion proceedings of the 14th ACM international conference on future energy systems, e-Energy \u201923 Companion. Association for Computing Machinery, New York, NY, USA, pp 26\u201331. https:\/\/doi.org\/10.1145\/3599733.3600247","DOI":"10.1145\/3599733.3600247"},{"issue":"8","key":"904_CR34","doi-asserted-by":"publisher","first-page":"3086","DOI":"10.1016\/j.enpol.2008.03.035","volume":"36","author":"F Sensfu\u00df","year":"2008","unstructured":"Sensfu\u00df F, Ragwitz M, Genoese M (2008) The merit-order effect: a detailed analysis of the price effect of renewable electricity generation on spot market prices in germany. Ener Policy 36(8):3086\u20133094. https:\/\/doi.org\/10.1016\/j.enpol.2008.03.035 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S0301421508001717)","journal-title":"Ener Policy"},{"key":"904_CR35","first-page":"307","volume-title":"Contributions to the theory of games, annals of mathematics studies","author":"LS Shapley","year":"1953","unstructured":"Shapley LS (1953) A value for n-person games. In: Kuhn HW, Tucker AW (eds) Contributions to the theory of games, annals of mathematics studies, vol II. Princeton University Press, Princeton, pp 307\u2013317"},{"key":"904_CR36","doi-asserted-by":"publisher","first-page":"202","DOI":"10.1016\/j.erss.2015.12.020","volume":"13","author":"BK Sovacool","year":"2016","unstructured":"Sovacool BK (2016) How long will it take? Conceptualizing the temporal dynamics of energy transitions. Energy Res Soc Sci 13:202\u2013215. https:\/\/doi.org\/10.1016\/j.erss.2015.12.020 (https:\/\/www.sciencedirect.com\/science\/article\/pii\/S2214629615300827)","journal-title":"Energy Res Soc Sci"},{"key":"904_CR37","doi-asserted-by":"publisher","unstructured":"Sundararajan M, Najmi A (2020) The many Shapley values for model explanation. https:\/\/doi.org\/10.48550\/arXiv.1908.08474. ArXiv:1908.08474 [cs]","DOI":"10.48550\/arXiv.1908.08474"},{"key":"904_CR38","doi-asserted-by":"crossref","unstructured":"Veale M, Zuiderveen\u00a0Borgesius F (2021) Demystifying the draft EU artificial intelligence act. https:\/\/papers.ssrn.com\/abstract=3896852","DOI":"10.31235\/osf.io\/38p5f"},{"key":"904_CR39","doi-asserted-by":"publisher","unstructured":"Wang J, Wiens J, Lundberg S (2021) Shapley flow: a graph-based approach to interpreting model predictions. https:\/\/doi.org\/10.48550\/arXiv.2010.14592. ArXiv:2010.14592 [cs]","DOI":"10.48550\/arXiv.2010.14592"},{"issue":"1","key":"904_CR40","doi-asserted-by":"publisher","first-page":"1161","DOI":"10.1109\/TVCG.2021.3114864","volume":"28","author":"A Zytek","year":"2022","unstructured":"Zytek A, Liu D, Vaithianathan R, Veeramachaneni K (2022) Sibyl: understanding and addressing the usability challenges of machine learning in high-stakes decision making. IEEE Trans Visual Comput Graphics 28(1):1161\u20131171. https:\/\/doi.org\/10.1109\/TVCG.2021.3114864 (https:\/\/ieeexplore.ieee.org\/document\/9552849)","journal-title":"IEEE Trans Visual Comput Graphics"}],"container-title":["KI - K\u00fcnstliche Intelligenz"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s13218-026-00904-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/article\/10.1007\/s13218-026-00904-4","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/link.springer.com\/content\/pdf\/10.1007\/s13218-026-00904-4.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,6,16]],"date-time":"2026-06-16T21:10:26Z","timestamp":1781644226000},"score":1,"resource":{"primary":{"URL":"https:\/\/link.springer.com\/10.1007\/s13218-026-00904-4"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2026,3]]},"references-count":40,"journal-issue":{"issue":"1","published-print":{"date-parts":[[2026,3]]}},"alternative-id":["904"],"URL":"https:\/\/doi.org\/10.1007\/s13218-026-00904-4","relation":{},"ISSN":["0933-1875","1610-1987"],"issn-type":[{"value":"0933-1875","type":"print"},{"value":"1610-1987","type":"electronic"}],"subject":[],"published":{"date-parts":[[2026,3]]},"assertion":[{"value":"5 September 2025","order":1,"name":"received","label":"Received","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"12 March 2026","order":2,"name":"accepted","label":"Accepted","group":{"name":"ArticleHistory","label":"Article History"}},{"value":"30 March 2026","order":3,"name":"first_online","label":"First Online","group":{"name":"ArticleHistory","label":"Article History"}}]}}