pith. sign in

arxiv: 2606.19118 · v1 · pith:ZMW43O2Gnew · submitted 2026-06-17 · 💻 cs.AI · cs.LG· econ.GN· q-fin.EC

Analysing drivers and interdependencies in European electricity markets using XAI

Antoine Pesenti , Aidan O'Sullivan This is my paper

Pith reviewed 2026-06-26 20:44 UTC · model grok-4.3

classification 💻 cs.AI cs.LGecon.GNq-fin.EC
keywords electricity pricesexplainable AISHAPrenewablesEuropean marketsgas pricesinterconnectionsprice formation
0
0 comments X

The pith

Deep neural networks paired with explainable AI show solar power shapes European electricity prices far beyond its share of generation, while gas prices and interconnections act as steady drivers.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper trains DNNs on data from 39 European bidding zones to forecast electricity prices and then applies SHAP values plus an extended SSHAP aggregation to measure how much each input contributes to those predictions. This decomposition identifies solar generation as having an outsized effect on prices relative to its actual output, gas prices as a dominant factor everywhere, and transmission links between zones as important for how prices move together. The authors further build a synthetic model of a single EU-wide market to examine price formation under full integration. The approach turns accurate but opaque forecasts into concrete statements about which physical and economic factors actually set prices in interconnected systems.

Core claim

By training DNNs on European electricity price data and extracting feature attributions via SHAP and the SSHAP framework, the analysis establishes that renewable sources especially solar play a disproportionately important role in price formation despite lower generation shares, gas prices remain a dominant and consistent driver across markets, and interconnections significantly shape price dynamics, while a synthetic EU-wide single-price market is constructed to explore the counterfactual of full integration.

What carries the argument

SHAP values and SSHAP aggregation applied to DNN models trained on electricity price data from 39 bidding zones.

If this is right

  • Gas prices act as a dominant and consistent driver across all examined electricity markets.
  • Interconnections significantly shape price dynamics and underscore the interdependence of European electricity systems.
  • A synthetic model of a fully integrated EU market with a single price enables exploration of counterfactual price outcomes.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • If the solar attribution holds under higher renewable penetration, price volatility in sunny zones may decline faster than conventional models expect.
  • The same XAI pipeline could be rerun on future data to test whether the identified drivers shift after major policy changes such as expanded carbon pricing.
  • High correlations among energy variables may still require separate validation experiments to confirm that attributions are not artifacts of multicollinearity.

Load-bearing premise

The DNNs capture the true underlying drivers of price formation rather than spurious correlations, and the SHAP and SSHAP attributions faithfully reflect the model's decision process without distortion from feature correlations or model misspecification.

What would settle it

Periods of high solar output where the models do not attribute large price reductions to solar, or where removing gas price inputs fails to change predictions in line with the reported attributions.

Figures

Figures reproduced from arXiv: 2606.19118 by Aidan O'Sullivan, Antoine Pesenti.

Figure 1
Figure 1. Figure 1: Dominant generation source map showing lack of geograph [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: 2023-2025 historical prices in €/MWh showing high diver￾sity between the 40 bidding zones Representing the 39 bidding zones with their price means and standard deviations (SD) scattered as in [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Hierarchical clustering dendrogram showing three small [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗
Figure 3
Figure 3. Figure 3: Mean versus standard deviation plot showing high diversity [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figure 5
Figure 5. Figure 5: Correlations with electricity price plot highlighting very [PITH_FULL_IMAGE:figures/full_fig_p005_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: EU SHAP beeswarm plot showing that gas price is the [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: EU SSHAP beeswarm plot showing that solar is the most [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: SSHAP neighbours importance map showing that most [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Dominant SSHAP super-feature map showing solar being [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Dominant super-feature map highlighting that German [PITH_FULL_IMAGE:figures/full_fig_p010_10.png] view at source ↗
read the original abstract

Electricity markets are inherently complex systems characterised by strong nonlinearities, high-dimensional interactions, and increasing interdependence across regions. While deep neural networks (DNNs) have demonstrated strong predictive capabilities for electricity prices, their lack of interpretability limits their usefulness for understanding the underlying drivers of price formation. This paper addresses this gap by combining DNN models with explainable artificial intelligence (XAI) techniques to analyse the determinants of electricity prices across 39 European bidding zones. We employ SHAP (SHapley Additive exPlanations) to quantify feature contributions and apply and extend SSHAP, an aggregation framework to improve interpretability in high-dimensional settings. The analysis identifies that renewable energy sources, particularly solar, play a disproportionately important role in price formation despite their lower share in total power generation. Gas prices remain a dominant and consistent driver across electricity markets, while interconnections significantly shape price dynamics, highlighting the strong interdependence of European electricity systems. In addition, a synthetic EU-wide electricity market is constructed to explore the counterfactual scenario of a fully integrated market with a single price.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

3 major / 2 minor

Summary. The manuscript trains deep neural networks on data from 39 European electricity bidding zones to predict prices, then applies SHAP and an extended SSHAP aggregation method to attribute feature importance. It claims that renewable sources (particularly solar) exert a disproportionately large influence on price formation relative to their generation share, that gas prices are a dominant and consistent driver, and that cross-border interconnections materially shape price dynamics. A synthetic EU-wide market is also constructed to examine the counterfactual of full integration under a single price.

Significance. If the attributions can be shown to be robust, the scale of the analysis across dozens of zones and the SSHAP extension could provide actionable insights for renewable integration and market-coupling policy. The work also demonstrates a practical workflow for combining high-capacity predictors with post-hoc explanation in a high-stakes domain.

major comments (3)
  1. [Abstract and Methods] Abstract and Methods: The central claim that solar 'plays a disproportionately important role in price formation' and that gas/interconnections are 'dominant drivers' interprets SHAP/SSHAP values as structural effects. No description is given of how the DNN architecture, training procedure, or SSHAP extension mitigates the known instability of additive attributions under multicollinearity (solar generation correlates with demand, irradiance, other renewables, and time-of-day).
  2. [Results] Results (feature-importance rankings): The reported importance orderings are presented without accompanying robustness diagnostics such as (i) retraining after orthogonalizing correlated renewable features, (ii) comparison against permutation importance or conditional SHAP variants, or (iii) any causal-identification strategy. This leaves open whether the 'disproportionate role' finding survives feature dependence.
  3. [Synthetic-market counterfactual] Synthetic-market counterfactual: The construction of the EU-wide single-price scenario relies on the same fitted models; without explicit checks that the DNNs generalize under the altered interconnection structure, the counterfactual price distribution may simply extrapolate the same correlational patterns rather than reflect a new equilibrium.
minor comments (2)
  1. [Methods] Notation for SSHAP aggregation is introduced at a high level; a concise mathematical definition (or pseudocode) in the Methods section would improve reproducibility.
  2. [Introduction] The manuscript would benefit from explicit comparison to earlier econometric studies of European electricity prices that already document gas-price and interconnection effects, to clarify the incremental contribution of the XAI layer.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. We address each major comment point by point below, indicating where revisions will strengthen the presentation of methods, results, and limitations.

read point-by-point responses

  1. Referee: Abstract and Methods: The central claim that solar 'plays a disproportionately important role in price formation' and that gas/interconnections are 'dominant drivers' interprets SHAP/SSHAP values as structural effects. No description is given of how the DNN architecture, training procedure, or SSHAP extension mitigates the known instability of additive attributions under multicollinearity (solar generation correlates with demand, irradiance, other renewables, and time-of-day).

    Authors: We acknowledge that multicollinearity can affect the stability of SHAP attributions, particularly for solar generation which correlates with demand and other variables. The DNN architecture captures nonlinear interactions, and the SSHAP aggregation across 39 zones is intended to improve robustness by averaging attributions. However, the original submission did not explicitly address mitigation strategies or compare to conditional variants. We will revise the Methods section to discuss this known limitation of additive feature attributions and add a supplementary comparison using permutation importance to evaluate ranking stability. revision: yes

  2. Referee: Results (feature-importance rankings): The reported importance orderings are presented without accompanying robustness diagnostics such as (i) retraining after orthogonalizing correlated renewable features, (ii) comparison against permutation importance or conditional SHAP variants, or (iii) any causal-identification strategy. This leaves open whether the 'disproportionate role' finding survives feature dependence.

    Authors: Our work applies XAI to observational predictive models and does not claim causal identification, which would require additional assumptions or instruments outside the paper's scope. Orthogonalization of features risks distorting the natural time-series structure of electricity data. We will add permutation importance rankings as a robustness check in the revised results and supplementary materials. We will also clarify in the text that the reported importance reflects model attributions under observed correlations rather than isolated structural effects. revision: partial

  3. Referee: Synthetic-market counterfactual: The construction of the EU-wide single-price scenario relies on the same fitted models; without explicit checks that the DNNs generalize under the altered interconnection structure, the counterfactual price distribution may simply extrapolate the same correlational patterns rather than reflect a new equilibrium.

    Authors: The synthetic EU-wide market is presented as an exploratory counterfactual to illustrate potential effects of full integration, using the trained models on adjusted interconnection features. We agree that explicit generalization checks under structural changes are absent and that results may reflect extrapolation. We will revise the relevant section to explicitly discuss this limitation, clarify the illustrative nature of the exercise, and note it as an area for future work involving equilibrium models or simulated structural shifts. revision: yes

Circularity Check

0 steps flagged

Empirical XAI attribution on observational data; no derivation reduces to inputs by construction

full rationale

The paper trains DNNs on European electricity market data and applies SHAP/SSHAP for feature attribution. No equations, first-principles derivations, or 'predictions' are shown that reduce to fitted parameters or self-citations by construction. The central claims are empirical interpretations of model outputs on external data. SSHAP is mentioned as an extension but is not load-bearing for the reported drivers; the analysis remains falsifiable against held-out market observations. This matches the default expectation of no significant circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only; no free parameters, axioms, or invented entities are described.

pith-pipeline@v0.9.1-grok · 5717 in / 1063 out tokens · 20829 ms · 2026-06-26T20:44:21.706553+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

42 extracted references · 30 canonical work pages · 2 internal anchors

  1. [1]

    URLhttps://eepublicdownloads.blob.core.windows.net/strapi-t est-assets/strapi-assets/entso-e_Market_report_2025.pdf

    ENTSO-E, ENTSO-E Market Report 2025 (2025). URLhttps://eepublicdownloads.blob.core.windows.net/strapi-t est-assets/strapi-assets/entso-e_Market_report_2025.pdf

    2025

  2. [2]

    URLhttps://iea.blob.core.windows.net/assets/b73798cb-e452-4 2b9-9d8a-07542de7a041/Electricity_2026.pdf

    IEA, Electricity 2026: Analysis and Forecast to 2030 (2026). URLhttps://iea.blob.core.windows.net/assets/b73798cb-e452-4 2b9-9d8a-07542de7a041/Electricity_2026.pdf

    2026

  3. [3]

    Pesenti, A

    A. Pesenti, A. O’Sullivan, Explaining deep neural network models for electricity price forecasting with XAI, Energy and AI 21 (2025) 100532.doi:10.1016/j.egyai.2025.100532. URLhttps://linkinghub.elsevier.com/retrieve/pii/S2666546825 000643

    work page doi:10.1016/j.egyai.2025.100532 2025

  4. [4]

    Maciejowska, B

    K. Maciejowska, B. Uniejewski, R. Weron, Forecasting Electricity Prices, arXiv:2204.11735 [q-fin] (Apr. 2022).doi:10.48550/arXiv.2 204.11735. URLhttp://arxiv.org/abs/2204.11735

    work page doi:10.48550/arxiv.2 2022

  5. [5]

    H. A. Gil, C. Gómez-Quiles, A. Gómez-Expósito, J. R. Santos, Forecasting Prices in Electricity Markets: Needs, Tools and Lim- itations, in: A. Sorokin, S. Rebennack, P. M. Pardalos, N. A. Il- iadis, M. V. F. Pereira (Eds.), Handbook of Networks in Power Systems I, Springer, Berlin, Heidelberg, 2012, pp. 123–150.doi: 10.1007/978-3-642-23193-3_5. URLhttps:/...

    work page doi:10.1007/978-3-642-23193-3_5 2012

  6. [6]

    Contreras, R

    J. Contreras, R. Espinola, F. Nogales, A. Conejo, ARIMA models to predict next-day electricity prices, IEEE Transactions on Power Systems 18 (3) (2003) 1014–1020, conference Name: IEEE Transac- tions on Power Systems.doi:10.1109/TPWRS.2002.804943. URLhttps://ieeexplore.ieee.org/abstract/document/1216141

    work page doi:10.1109/tpwrs.2002.804943 2003

  7. [7]

    A. J. Conejo, J. Contreras, R. Espínola, M. A. Plazas, Forecasting electricity prices for a day-ahead pool-based electric energy market, International Journal of Forecasting 21 (3) (2005) 435–462.doi: 10.1016/j.ijforecast.2004.12.005. URLhttps://www.sciencedirect.com/science/article/pii/S01692 07004001311

    work page doi:10.1016/j.ijforecast.2004.12.005 2005

  8. [8]

    Würzburg, X

    K. Würzburg, X. Labandeira, P. Linares, Renewable generation and electricity prices: Taking stock and new evidence for Germany and Austria, Energy Economics 40 (2013) S159–S171.doi:10.1016/j.en eco.2013.09.011. URLhttps://www.sciencedirect.com/science/article/pii/S01409 88313002065

    work page doi:10.1016/j.en 2013

  9. [9]

    Csereklyei, S

    Z. Csereklyei, S. Qu, T. Ancev, The effect of wind and solar power generationonwholesaleelectricitypricesinAustralia, EnergyPolicy 131 (2019) 358–369.doi:10.1016/j.enpol.2019.04.007. URLhttps://www.sciencedirect.com/science/article/pii/S03014 21519302435

    work page doi:10.1016/j.enpol.2019.04.007 2019

  10. [10]

    J. Seel, A. D. Mills, R. H. Wiser, Impacts of High Variable Re- newable Energy Futures on Wholesale Electricity Prices, and on Electric-Sector Decision Making (May 2018). URLhttps://escholarship.org/uc/item/2xq5d6c9

    2018

  11. [11]

    B. Acar, O. Selcuk, S. A. Dastan, The merit order effect of wind and river type hydroelectricity generation on Turkish electricity prices, Energy Policy 132 (2019) 1298–1319.doi:10.1016/j.enpol.2019.0 7.006. URLhttps://www.sciencedirect.com/science/article/pii/S03014 21519304483

    work page doi:10.1016/j.enpol.2019.0 2019

  12. [12]

    L. Wen, K. Suomalainen, B. Sharp, M. Yi, M. S. Sheng, Impact of wind-hydro dynamics on electricity price: A seasonal spatial econo- metric analysis, Energy 238 (2022) 122076.doi:10.1016/j.energy.2 021.122076. URLhttps://www.sciencedirect.com/science/article/pii/S03605 44221023240

    work page doi:10.1016/j.energy.2 2022

  13. [13]

    Spectroscopic investigation of the tungsten deuteride sputtering in the EAST divertor

    T. Stringer, M. Joanis, S. Abdoli, Power generation mix and elec- tricity price, Renewable Energy 221 (2024) 119761.doi:10.1016/j. renene.2023.119761. URLhttps://www.sciencedirect.com/science/article/pii/S09601 48123016762

    work page doi:10.1016/j 2024

  14. [14]

    S. K. Aggarwal, L. M. Saini, A. Kumar, Electricity price forecast- ing in deregulated markets: A review and evaluation, International Journal of Electrical Power & Energy Systems 31 (1) (2009) 13–22. doi:10.1016/j.ijepes.2008.09.003. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0142061508 000884

    work page doi:10.1016/j.ijepes.2008.09.003 2009

  15. [15]

    R. Weron, Electricity price forecasting: A review of the state-of-the- art with a look into the future, International Journal of Forecasting 30 (4) (2014) 1030–1081.doi:10.1016/j.ijforecast.2014.08.008. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0169207014 001083

    work page doi:10.1016/j.ijforecast.2014.08.008 2014

  16. [16]

    V.Laitsos, G.Vontzos, P.Paraschoudis, E.Tsampasis, D.Bargiotas, L. H. Tsoukalas, The State of the Art Electricity Load and Price Forecasting for the Modern Wholesale Electricity Market, Energies 17 (22) (2024) 5797, number: 22.doi:10.3390/en17225797. URLhttps://www.mdpi.com/1996-1073/17/22/5797

    work page doi:10.3390/en17225797 2024

  17. [17]

    Zakeri, I

    B. Zakeri, I. Staffell, P. E. Dodds, M. Grubb, P. Ekins, J. Jääskeläi- nen, S. Cross, K. Helin, G. Castagneto Gissey, The role of natural gas in setting electricity prices in Europe, Energy Reports 10 (2023) 2778–2792.doi:10.1016/j.egyr.2023.09.069. URLhttps://www.sciencedirect.com/science/article/pii/S23524 84723013057

    work page doi:10.1016/j.egyr.2023.09.069 2023

  18. [18]

    Khosravi, S

    A. Khosravi, S. Nahavandi, D. Creighton, Quantifying uncertainties of neural network-based electricity price forecasts, Applied Energy 112 (2013) 120–129.doi:10.1016/j.apenergy.2013.05.075. URLhttps://www.sciencedirect.com/science/article/pii/S03062 61913004881

    work page doi:10.1016/j.apenergy.2013.05.075 2013

  19. [19]

    J. Lago, F. De Ridder, B. De Schutter, Forecasting spot elec- tricity prices: Deep learning approaches and empirical compari- son of traditional algorithms, Applied Energy 221 (2018) 386–405. doi:10.1016/j.apenergy.2018.02.069. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0306261918 30196X

    work page doi:10.1016/j.apenergy.2018.02.069 2018

  20. [20]

    Doshi-Velez, B

    F. Doshi-Velez, B. Kim, Towards A Rigorous Science of Inter- pretable Machine Learning (Feb. 2017). URLhttps://arxiv.org/abs/1702.08608v2

    Pith/arXiv arXiv 2017

  21. [21]

    Bodria, F

    F. Bodria, F. Giannotti, R. Guidotti, F. Naretto, D. Pedreschi, S. Rinzivillo, Benchmarking and Survey of Explanation Methods for Black Box Models, arXiv:2102.13076 [cs] (Feb. 2021). URLhttp://arxiv.org/abs/2102.13076

    arXiv 2021

  22. [22]

    Molnar, Interpretable Machine Learning, Molnar, 2022

    C. Molnar, Interpretable Machine Learning, Molnar, 2022. URLhttps://christophm.github.io/interpretable-ml-book/

    2022

  23. [23]

    Machlev, L

    R. Machlev, L. Heistrene, M. Perl, K. Levy, J. Belikov, S. Mannor, Y. Levron, Explainable Artificial Intelligence (XAI) techniques for energy and power systems: Review, challenges and opportunities, Energy and AI 9 (2022) 100169.doi:10.1016/j.egyai.2022.100169. URLhttps://linkinghub.elsevier.com/retrieve/pii/S2666546822 000246

    work page doi:10.1016/j.egyai.2022.100169 2022

  24. [24]

    Lundberg, S.-I

    S. Lundberg, S.-I. Lee, A Unified Approach to Interpreting Model Predictions, arXiv:1705.07874 [cs, stat] (Nov. 2017). URLhttp://arxiv.org/abs/1705.07874

    Pith/arXiv arXiv 2017

  25. [25]

    URL https://doi

    M. T. Ribeiro, S. Singh, C. Guestrin, "Why Should I Trust You?": Explaining the Predictions of Any Classifier, in: Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Dis- covery and Data Mining, ACM, San Francisco California USA, 2016, pp. 1135–1144.doi:10.1145/2939672.2939778. URLhttps://dl.acm.org/doi/10.1145/2939672.2939778

    work page doi:10.1145/2939672.2939778 2016

  26. [26]

    Heistrene, R

    L. Heistrene, R. Machlev, M. Perl, J. Belikov, D. Baimel, K. Levy, S. Mannor, Y. Levron, Explainability-based Trust Algorithm for electricity price forecasting models, Energy and AI 14 (2023) 100259.doi:10.1016/j.egyai.2023.100259. URLhttps://linkinghub.elsevier.com/retrieve/pii/S2666546823 000319

    work page doi:10.1016/j.egyai.2023.100259 2023

  27. [27]

    Heistrene, J

    L. Heistrene, J. Belikov, D. Baimel, L. Katzir, R. Machlev, K. Levy, S. Mannor, Y. Levron, An Improved and Explainable Electric- ity Price Forecasting Model via SHAP-based Error Compensation Approach, IEEE Transactions on Artificial Intelligence (2024) 1– 11Conference Name: IEEE Transactions on Artificial Intelligence. 10 doi:10.1109/TAI.2024.3455313. URL...

    work page doi:10.1109/tai.2024.3455313 2024

  28. [28]

    Random forests,

    L. Breiman, Random Forests, Machine Learning 45 (1) (2001) 5–32. doi:10.1023/A:1010933404324. URLhttps://doi.org/10.1023/A:1010933404324

    work page doi:10.1023/a:1010933404324 2001

  29. [29]

    Fisher, C

    A. Fisher, C. Rudin, F. Dominici, All Models are Wrong, but Many are Useful: Learning a Variable’s Importance by Studying an En- tire Class of Prediction Models Simultaneously, Journal of Machine Learning Research 20 (177) (2019) 1–81. URLhttp://jmlr.org/papers/v20/18-760.html

    2019

  30. [30]

    Trebbien, L

    J. Trebbien, L. Rydin Gorjão, A. Praktiknjo, B. Schäfer, D. Wit- thaut, Understanding electricity prices beyond the merit order principle using explainable AI, Energy and AI 13 (2023) 100250. doi:10.1016/j.egyai.2023.100250. URLhttps://linkinghub.elsevier.com/retrieve/pii/S2666546823 000228

    work page doi:10.1016/j.egyai.2023.100250 2023

  31. [31]

    S. M. Lundberg, G. G. Erion, S.-I. Lee, Consistent Individualized Feature Attribution for Tree Ensembles, arXiv:1802.03888 [cs] (Mar. 2019).doi:10.48550/arXiv.1802.03888. URLhttp://arxiv.org/abs/1802.03888

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1802.03888 2019

  32. [32]

    Tschora, E

    L. Tschora, E. Pierre, M. Plantevit, C. Robardet, Electricity price forecasting on the day-ahead market using machine learning, Ap- plied Energy 313 (2022) 118752.doi:10.1016/j.apenergy.2022.1187 52. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0306261922 002057

    work page doi:10.1016/j.apenergy.2022.1187 2022

  33. [33]

    M. M. Mascarenhas, M. Amelin, H. Kazmi, Bridging Accuracy and Explainability in Electricity Price Forecasting, in: 2024 20th Inter- national Conference on the European Energy Market (EEM), 2024, pp. 1–6.doi:10.1109/EEM60825.2024.10608857. URLhttps://ieeexplore.ieee.org/document/10608857/?arnumber= 10608857

    work page doi:10.1109/eem60825.2024.10608857 2024

  34. [34]

    J. Lago, G. Marcjasz, B. De Schutter, R. Weron, Forecasting day- aheadelectricityprices: Areviewofstate-of-the-artalgorithms, best practices and an open-access benchmark, Applied Energy 293 (2021) 116983.doi:10.1016/j.apenergy.2021.116983. URLhttps://linkinghub.elsevier.com/retrieve/pii/S0306261921 004529

    work page doi:10.1016/j.apenergy.2021.116983 2021

  35. [35]

    NEMOs, EUPHEMIA Public Description - Single Price Coupling Algorithm (Aug. 2024). URLhttps://www.nordpoolgroup.com/globalassets/download-cente r/single-day-ahead-coupling/euphemia-public-description.pdf

    2024

  36. [36]

    Aščerić, M

    A. Aščerić, M. Pantoš, M. Čepin, Correlation of Day-Ahead Elec- tric Energy Market Price with Renewable Energy Sources Genera- tion and Load Forecast, in: Advanced Technologies, Systems, and Applications VI, Springer Nature, Switzerland, 2022, pp. 99–109. doi:10.1007/978-3-030-90055-7_8

    work page doi:10.1007/978-3-030-90055-7_8 2022

  37. [37]

    Gasparella, D

    A. Gasparella, D. Koolen, A. Zucker, The Merit Order and Price- Setting Dynamics in European Electricity Markets (2023). URLhttps://publications.jrc.ec.europa.eu/repository/handle/ JRC134300

    2023

  38. [38]

    D. P. Kingma, J. Ba, Adam: A Method for Stochastic Optimization, arXiv:1412.6980 [cs] (Jan. 2017).doi:10.48550/arXiv.1412.6980. URLhttp://arxiv.org/abs/1412.6980

    work page internal anchor Pith review Pith/arXiv arXiv doi:10.48550/arxiv.1412.6980 2017

  39. [39]

    H. Chen, J. D. Janizek, S. Lundberg, S.-I. Lee, True to the Model or True to the Data?, arXiv:2006.16234 [cs, stat] (Jun. 2020). URLhttp://arxiv.org/abs/2006.16234

    arXiv 2006

  40. [40]

    H. Chen, I. C. Covert, S. M. Lundberg, S.-I. Lee, Algorithms to estimate Shapley value feature attributions, Nat Mach Intell 5 (6) (2023) 590–601.doi:10.1038/s42256-023-00657-x. URLhttps://www.nature.com/articles/s42256-023-00657-x

    work page doi:10.1038/s42256-023-00657-x 2023

  41. [41]

    Štrumbelj, I

    E. Štrumbelj, I. Kononenko, Explaining prediction models and indi- vidual predictions with feature contributions, Knowl Inf Syst 41 (3) (2014) 647–665.doi:10.1007/s10115-013-0679-x. URLhttp://link.springer.com/10.1007/s10115-013-0679-x

    work page doi:10.1007/s10115-013-0679-x 2014

  42. [42]

    L. S. Shapley, A value for n-person games (1953), in: A. E. Roth (Ed.), The Shapley Value: Essays in Honor of Lloyd S. Shapley, Cambridge University Press, Cambridge, 1988, pp. 31–40.doi:10.1 017/CBO9780511528446.003. URLhttps://www.cambridge.org/core/books/shapley-value/value -for-nperson-games/1AA9D343DE7A87A97F69E999D329B57A 11

    1953