1. Proceedings
  2. I3M
  3. 2022
  4. EMSS
  5. 10.46354/i3m.2022.emss.042

Utilizing Interpretable Machine Learning to detect Dynamics in Energy Communities 

  • a Jan Zenisek ,
  • b Florian Bachinger, 
  • c Erik Pitzer, 
  • d Stefan Wagner, 
  • e Michael Affenzeller 
  • a,b,c,d,e  Heuristic and Evolutionary Algorithms Laboratory, University of Applied Sciences Upper Austria, Softwarepark 11, Hagenberg, 4232, Austria 
  • a,eInstitute for Symbolic Artificial Intelligence, Johannes Kepler University, Altenberger Straße 69, Linz, 4040, Austria
Cite as
Zenisek J., Bachinger F., Pitzer E., Wagner S., and Affenzeller M. (2022).,Utilizing Interpretable Machine Learning to detect Dynamics in Energy Communities. Proceedings of the 34th European Modeling & Simulation Symposium (EMSS 2022). , 042 . DOI: https://doi.org/10.46354/i3m.2022.emss.042

Abstract

With the growing use of machine learning models in many critical domains, research regarding making these models, as well as their predictions, more explainable has intensified in the last few years. In this paper, we present extensions to the machine learning based data mining technique Variable Interaction Networks (VIN), to integrate existing domain knowledge and thus, enable more meaningful analysis. Several tests on data from a case study concerned with long-term monitored photovoltaic systems, verify the feasibility of our approach to provide valuable, human-interpretable insights. In particular, we show the successful application of root-cause detection in scenarios with changing system conditions.

References

  1. Affenzeller, M., Winkler, S. M., Kronberger, G., Kom menda, M., Burlacu, B., and Wagner, S. (2014). Gain ing deeper insights in symbolic regression. In Ge netic Programming Theory and Practice XI, pages 175–190. Springer.
  2. Breiman, L. (2001). Random forests. Machine learning, 45(1):5–32.
  3. Draper, N. and Smith, H. (1966). Applied linear regression. Faria, P., Barreto, R., and Vale, Z. (2019). Demand response in energy communities considering the share of photo voltaic generation from public buildings. In 2019 Inter national Conference on Smart Energy Systems and Tech nologies (SEST), pages 1–6.
  4. Fisher, A., Rudin, C., and Dominici, F. (2019). All models are wrong, but many are useful: Learning a variable’s importance by studying an entire class of prediction models simultaneously. J. Mach. Learn. Res., 20(177):1– 81.
  5. Friedman, J. H. and Popescu, B. E. (2008). Predictive learning via rule ensembles. The annals of applied statistics, pages 916–954. 
  6. Haiden, T., Kann, A., Pistotnik, G., Stadlbacher, K., and Wittmann, C. (2009). Integrated nowcasting through comprehensive analysis (inca)—system description. ZAMG Rep, 61.
  7. Hooker, G. (2004). Discovering additive structure in black box functions. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining, pages 575–580.
  8. Kommenda, M., Kronberger, G., Feilmayr, C., and Affen zeller, M. (2011). Data mining using unguided symbolic regression on a blast furnace dataset. In European Con ference on the Applications of Evolutionary Computation, pages 274–283. Springer.
  9. Koza, J. R. (1994). Genetic programming as a means for programming computers by natural selection. Statistics and computing, 4(2):87–112.
  10. Kronberger, G., Burlacu, B., Kommenda, M., Winkler, S., and Affenzeller, M. (2017). Measures for the evaluation and comparison of graphical model structures. Lecture Notes in Computer Science, 10671:283–290.
  11. Kronberger, G., Fink, S., Kommenda, M., and Affenzeller, M. (2011). Macro-economic time series modeling and interaction networks. In European Conference on the Applications of Evolutionary Computation, pages 101–110. Springer. 
  12. Lundberg, S. M. and Lee, S.-I. (2017). A unified approach to interpreting model predictions. Advances in neural information processing systems, 30. 
  13. Molnar, C. (2022). Interpretable Machine Learning: A Guide for Making Black Box Models Explainable. Accessed: May 19, 2022. [Online]. Available: https://christophm.github.io/interpretable-ml 
    book/. 
  14. Shapley, L. S. (1953). A value for n-person games. In: Kuhn, H. and Tucker, A., Eds., Contributions to the Theory of Games II, 2.28:307–317.
  15. Winkler, S. M., Kronberger, G., Affenzeller, M., and Stekel, H. (2013). Variable interaction networks in medical data. International Journal of Privacy and Health Information Management (IJPHIM), 1(2):1–16.
  16. Winkler, S. M., Kronberger, G., Kommenda, M., Fink, S., and Affenzeller, M. (2015). Dynamics of predictability and variable influences identified in financial data using sliding window machine learning. In International Con ference on Computer Aided Systems Theory, pages 326– 333. Springer. 
  17. Zenisek, J., Kronberger, G., Wolfartsberger, J., Wild, N., and Affenzeller, M. (2020). Concept drift detection with variable interaction networks. Lecture Notes in Computer Science, 12013:296–303. 
  18. Zenisek, J., Wolfartsberger, J., Sievi, C., and Affenzeller, M. (2018). Streaming synthetic time series for simulated condition monitoring. IFAC-PapersOnLine, 51(11):643– 648.