Contextual Background Estimation for Explainable AI in Temperature Prediction
Source abstract
Abstract
Accurate weather prediction and electrical load modeling are critical for optimizing energy systems and mitigating environmental impacts. This study explores the integration of the novel Mean Background Method and Background Estimation Method with Explainable Artificial Intelligence (XAI) with the aim to enhance the evaluation and understanding of time-series models in these domains. The electrical load or temperature predictions are regression-based problems. Some XAI methods, such as SHAP, require using the base value of the model as the background to provide an explanation. However, in contextualized situations, the default base value is not always the best choice. The selection of the background can significantly affect the corresponding Shapley values. This paper presents two innovative XAI methods designed to provide robust context-aware explanations for regression and time-series problems, addressing critical gaps in model interpretability. They can be used to improve background selection to make more conscious decisions and improve the understanding of predictions made by models that use time-series data.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Energy Load and Power Forecasting · Meteorological Phenomena and Simulations · Hydrological Forecasting Using AI
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Faculty of Science and Technology, University of Silesia in Katowice, Bedzinska 39, 41-200 Sosnowiec, Poland
Institute of Computer Science, Faculty of Science and Technology, University of Silesia in Katowice, Bedzinska 39, 41-200 Sosnowiec, Poland
Independent Researcher, 40-069 Katowice, Poland
References (56)
Show source-provided references
1. Moreno Escobar, J.J., Morales Matamoros, O., Tejeida Padilla, R., Lina Reyes, I., and Quintana Espinosa, H. (2021). A comprehensive review on smart grids: Challenges and opportunities. Sensors, 21.10.3390/s21216978 ↗
2. Strielkowski, W., Civín, L., Tarkhanova, E., Tvaronavičienė, M., and Petrenko, Y. (2021). Renewable energy in the sustainable development of electrical power sector: A review. Energies, 14.10.3390/en14248240 ↗
3. Energy and sustainable development in smart cities: An overview10.3390/smartcities5040071 ↗
4. Rahman, M.M., Dadon, S.H., He, M., Giesselmann, M., and Hasan, M.M. (2024). An Overview of Power System Flexibility: High Renewable Energy Penetration Scenarios. Energies, 17.10.3390/en17246393 ↗
5. Yu, K., Wei, Q., Xu, C., Xiang, X., and Yu, H. (2024). Distributed Low-Carbon Energy Management of Urban Campus for Renewable Energy Consumption. Energies, 17.10.3390/en17236182 ↗
6. AI-based solutions for grid stability and efficiency: Challenges, limitations, and opportunities
7. Deep learning in power systems research: A review
8. Deep learning methods and applications for electrical power systems: A comprehensive review10.1002/er.5331 ↗
9. Forootan, M.M., Larki, I., Zahedi, R., and Ahmadi, A. (2022). Machine learning and deep learning in energy systems: A review. Sustainability, 14.10.3390/su14084832 ↗
10. Enhancing smart grid electricity prediction with the fusion of intelligent modeling and XAI integration10.21833/ijaas.2024.05.025 ↗
11. Park, J., and Kang, D. (2024). Artificial Intelligence and Smart Technologies in Safety Management: A Comprehensive Analysis Across Multiple Industries. Appl. Sci., 14.10.3390/app142411934 ↗
12. Elmousalami, H., A Alnaser, A., and Kin Peng Hui, F. (2024). Advancing Smart Zero-Carbon Cities: High-Resolution Wind Energy Forecasting to 36 Hours Ahead. Appl. Sci., 14.10.3390/app142411918 ↗
13. Identifying drivers and mitigators for congestion and redispatch in the German electric power system with explainable AI10.1016/j.apenergy.2023.122351 ↗
14. Adaptive classifier ensemble for multibiometric Verification10.1016/j.procs.2024.09.242 ↗
15. Hamrani, A., Medarametla, A., John, D., and Agarwal, A. (2024). Machine-Learning-Driven Optimization of Cold Spray Process Parameters: Robust Inverse Analysis for Higher Deposition Efficiency. Coatings, 15.10.3390/coatings15010012 ↗
16. Ali, A., Faheem, Z.B., Waseem, M., Draz, U., Safdar, Z., Hussain, S., and Yaseen, S. (2020, January 5–7). Systematic review: A state of art ML based clustering algorithms for data mining. Proceedings of the 2020 IEEE 23rd International Multitopic Conference (INMIC), Bahawalpur, Pakistan.10.1109/INMIC50486.2020.9318060 ↗
17. A preliminary study on the dispersed classification system for recognizing safety of drivers’ maneuvers10.1016/j.procs.2023.10.252 ↗
18. Machine learning advances for time series forecasting10.1111/joes.12429 ↗
19. Wrobel, K., Doroz, R., Porwik, P., Orczyk, T., Cavalcante, A.B., and Grajzer, M. (2022, January 28–30). Features of Hand-Drawn Spirals for Recognition of Parkinson’s Disease. Proceedings of the Asian Conference on Intelligent Information and Database Systems, Ho Chi Minh City, Vietnam.10.1007/978-3-031-21967-2_37 ↗
20. Explainable Artificial Intelligence (XAI) techniques for energy and power systems: Review, challenges and opportunities10.1016/j.egyai.2022.100169 ↗
21. An explainable AI framework for robust and transparent data-driven wind turbine power curve models10.1016/j.egyai.2023.100328 ↗
22. Panagoulias, D.P., Sarmas, E., Marinakis, V., Virvou, M., Tsihrintzis, G.A., and Doukas, H. (2023). Intelligent decision support for energy management: A methodology for tailored explainability of artificial intelligence analytics. Electronics, 12.10.3390/electronics12214430 ↗
23. Davydko, O., Pavlov, V., and Longo, L. (2023, January 26–28). Selecting Textural Characteristics of Chest X-Rays for Pneumonia Lesions Classification with the Integrated Gradients XAI Attribution Method. Proceedings of the World Conference on Explainable Artificial Intelligence, Lisbon, Portugal.10.1007/978-3-031-44064-9_36 ↗
24. Mahendran, A., and Vedaldi, A. (2016, January 11–14). Salient deconvolutional networks. Proceedings of the Computer Vision—ECCV 2016: 14th European Conference, Amsterdam, The Netherlands. Proceedings, Part VI 14.10.1007/978-3-319-46466-4_8 ↗
25. Victoria, A.H., Tiwari, R.S., and Ghulam, A.K. (2024). Libraries for Explainable Artificial Intelligence (EXAI): Python. Explainable AI (XAI) for Sustainable Development, Chapman and Hall/CRC.10.1201/9781003457176-13 ↗
26. Lundberg, S. (2017). A unified approach to interpreting model predictions. arXiv.
27. Ribeiro, M.T., Singh, S., and Guestrin, C. (2016, January 13–17). “Why should i trust you?” Explaining the predictions of any classifier. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, San Francisco, CA, USA.10.1145/2939672.2939778 ↗
28. Monteiro, W.R., and Reynoso-Meza, G. (2022). On the generation of global surrogate models through unconstrained multi-objective optimization. arXiv.10.21203/rs.3.rs-1409615/v1 ↗
29. Random forests10.1023/A:1010933404324 ↗
30. Greedy function approximation: A gradient boosting machine10.1214/aos/1013203451 ↗
31. Chambers, J.M. (2017). Linear models. Statistical Models in S, Routledge.10.1201/9780203738535 ↗
32. Decision trees10.1002/wics.1278 ↗
33. Intrinsic and post-hoc XAI approaches for fingerprint identification and response prediction in smart manufacturing processes10.1007/s10845-023-02266-2 ↗
34. Selvaraju, R.R., Cogswell, M., Das, A., Vedantam, R., Parikh, D., and Batra, D. (2017, January 22–29). Grad-cam: Visual explanations from deep networks via gradient-based localization. Proceedings of the IEEE International Conference on Computer Vision, Venice, Italy.10.1109/ICCV.2017.74 ↗
35. An efficient explanation of individual classifications using game theory
36. Yaprakdal, F., and Varol Arısoy, M. (2023). A multivariate time series analysis of electrical load forecasting based on a hybrid feature selection approach and explainable deep learning. Appl. Sci., 13.10.3390/app132312946 ↗
37. Powroźnik, P., and Szcześniak, P. (2024). Predictive Analytics for Energy Efficiency: Leveraging Machine Learning to Optimize Household Energy Consumption. Energies, 17.10.3390/en17235866 ↗
38. Laitsos, V., Vontzos, G., Paraschoudis, P., Tsampasis, E., Bargiotas, D., and Tsoukalas, L.H. (2024). The State of the Art Electricity Load and Price Forecasting for the Modern Wholesale Electricity Market. Energies, 17.10.20944/preprints202411.0165.v1 ↗
39. Introducing explainability in sequence-to-sequence learning for short-term load forecasting10.1016/j.epsr.2022.108366 ↗
40. Grzeszczyk, T.A., and Grzeszczyk, M.K. (2022). Justifying short-term load forecasts obtained with the use of neural models. Energies, 15.10.3390/en15051852 ↗
41. Enhancing smart grid load forecasting: An attention-based deep learning model integrated with federated learning and XAI for security and interpretability
42. Abumohsen, M., Owda, A.Y., and Owda, M. (2023). Electrical load forecasting using LSTM, GRU, and RNN algorithms. Energies, 16.10.3390/en16052283 ↗
43. Cordeiro-Costas, M., Villanueva, D., Eguía-Oller, P., Martínez-Comesaña, M., and Ramos, S. (2023). Load forecasting with machine learning and deep learning methods. Appl. Sci., 13.10.3390/app13137933 ↗
44. Probabilistic electric load forecasting: A tutorial review10.1016/j.ijforecast.2015.11.011 ↗
45. Ungureanu, S., Topa, V., and Cziker, A.C. (2021). Analysis for non-residential short-term load forecasting using machine learning and statistical methods with financial impact on the power market. Energies, 14.10.3390/en14216966 ↗
46. The impact of social and weather drivers on the historical electricity demand in Europe10.1016/j.apenergy.2018.07.108 ↗
47. Explainability and interpretability in electric load forecasting using machine learning techniques—A review10.1016/j.egyai.2024.100358 ↗
48. A comparative study of methods for estimating model-agnostic Shapley value explanations10.1007/s10618-024-01016-z ↗
49. Estimating Quality of Approximated Shapley Values Using Conformal Prediction
50. Yuan, H., Liu, M., Kang, L., Miao, C., and Wu, Y. (2022). An empirical study of the effect of background data size on the stability of SHapley Additive exPlanations (SHAP) for deep learning models. arXiv.
51. Toward explainable artificial intelligence for regression models: A methodological perspective10.1109/MSP.2022.3153277 ↗
52. Exploring how independent variables influence parking occupancy prediction: Toward a model results explanation with SHAP values10.1007/s13748-022-00291-5 ↗
53. Książek, W. (2024). Explainable Thyroid Cancer Diagnosis Through Two-Level Machine Learning Optimization with an Improved Naked Mole-Rat Algorithm. Cancers, 16.10.3390/cancers16244128 ↗
54. Bento, J., Saleiro, P., Cruz, A.F., Figueiredo, M.A., and Bizarro, P. (2021, January 14–18). Timeshap: Explaining recurrent models through sequence perturbations. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery & Data Mining, Singapore.10.1145/3447548.3467166 ↗
55. (2024, March 01). Max Planck Institute for Biogeochemistry. Jena, Germany. 2009–2016. Available online: https://www.bgc-jena.mpg.de/wetter/.
56. (2024, March 01). Instytut Meteorologii i Gospodarki Wodnej Państwowy Instytut Badawczy. 2015–2023. Available online: https://danepubliczne.imgw.pl/.