Hyperparameteroptimierung für Lastprognose/Efficient hyperparameter optimization for forecasts

Inhaltsverzeichnis

Bibliographische Infos

Cover der Ausgabe: wt Werkstattstechnik online Jahrgang 116 (2026), Heft 04
Open Access Vollzugriff

wt Werkstattstechnik online

Jahrgang 116 (2026), Heft 04

Autor:innen:
Verlag
VDI fachmedien, Düsseldorf
Copyrightjahr
2026
ISSN-Online
1436-4980
ISSN-Print
1436-4980

Kapitelinformationen

Open Access Vollzugriff

Jahrgang 116 (2026), Heft 04

Hyperparameteroptimierung für Lastprognose/Efficient hyperparameter optimization for forecasts

Autor:innen:
ISSN-Print
1436-4980
ISSN-Online
1436-4980
Kapitelvorschau:

Die Studie untersucht, wie ressourcenschonende Strategien zur Hyperparameteroptimierung die Genauigkeit und die Laufzeit industrieller Lastprognosen beeinflussen. Mit einem Taguchi-Design wurden zwei Modelle des maschinellen Lernens mit verschiedenen vereinfachenden Verfahren, sogenannten Pruning- und Subsampling-Methoden, getestet. Zufälliges Subsampling auf 30 % der Daten und Hyperband-Pruning erzielten teils bessere Prognosen bei deutlich geringerem Rechenaufwand.

Literaturverzeichnis

  1. [1] Paulus, M.; Borggrefe, F.: The potential of demand-side management in energy-intensive industries for electricity markets in Germany. Applied Energy 88 (2011) 2, pp.432–441, doi.org/10.1016/j.apenergy.2010.03.017 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  2. [2] Sharma, M.; Mittal, N.; Mishra, A. et al.: Survey of Electricity Demand Forecasting and Demand Side Management Techniques in Different Sectors to Identify Scope for Improvement. Smart Grids and Sustainable Energy 8 (2023) 2, #9, doi.org/10.1007/s40866–023–00168-z Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  3. [3] Hutter, F.; Kotthoff, L.; Vanschoren, J. (eds.): Automated Machine Learning: Methods, Systems, Challenges. Cham: Springer International Publishing 2019, doi.org/10.1007/978–3–030–05318–5 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  4. [4] Probst, P.; Wright, M. N.; Boulesteix, A.: Hyperparameters and tuning strategies for random forest. WIREs Data Mining and Knowledge Discovery 9 (2019) 3, pp.e1301, doi.org/10.1002/widm.1301 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  5. [5] Diaz, G. I.; Fokoue-Nkoutche, A.; Nannicini, G.; Samulowitz, H.: An effective algorithm for hyperparameter optimization of neural networks. IBM Journal of Research and Development 61 (2017) 4/5, pp. 9:1–9:11, doi.org/10.1147/JRD.2017.2709578 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  6. [6] Yang, L.; Shami, A.: On hyperparameter optimization of machine learning algorithms: Theory and practice. Neurocomputing 415 (2020), pp. 295–316, doi.org/10.1016/j.neucom.2020.07.061 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  7. [7] Andonie, R.: Hyperparameter optimization in learning systems. Journal of Membrane Computing 1 (2019) 4, pp. 279–291, https://doi.org/10.1007/s41965–019–00023–0 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  8. [8] Egele, R.; Guyon, I.; Sun, Y.; Balaprakash, P.: Is One Epoch All You Need For Multi-Fidelity Hyperparameter Optimization? ArXiv 2023, https://doi.org/10.48550/arXiv.2307.15422 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  9. [9] Wu, J.; Chen, X.-Y.; Zhang, H.; Xiong, L.-D.; Lei, H.; Deng, S.-H.: Hyperparameter Optimization for Machine Learning Models Based on Bayesian Optimization. Journal of Electronic Science and Technology 17, (2019) 1, pp. 26–40, http://doi.org/10.11989/JEST.1674–862X.80904120 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  10. [10] Puentes G., D. E.; Barrios H., C. J.; Navaux, P. O. A.: Hyperparameter Optimization for Convolutional Neural Networks with Genetic Algorithms and Bayesian Optimization. 2022 IEEE Latin American Conference on Computational Intelligence (LA-CCI). Montevideo, Uruguay: IEEE 2022, pp. 1–5, https://doi.org/10.1109/LA-CCI54402.2022.9981104 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  11. [11] Li, L.; Jamieson, K.; Rostamizadeh, A. et al.: A System for Massively Parallel Hyperparameter Tuning. ArXiv 20218, https://doi.org/10.48550/arXiv.1810.05934 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  12. [12] Li, L.; Jamieson, K.; DeSalvo, G. et al.: Hyperband: A Novel Bandit-Based Approach to Hyperparameter Optimization. ArXiv 2018, https://doi.org/10.48550/arXiv.1603.06560 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  13. [13] Mendes, P.; Casimiro, M.; Romano, P. et al.: TrimTuner: Efficient Optimization of Machine Learning Jobs in the Cloud via Sub-Sampling. 28th International Symposium on Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS). Nice, France, IEEE 2020, pp. 1–8, https://doi.org/10.1109/MASCOTS50786.2020.9285971 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  14. [14] Klein, A.; Falkner, S.; Bartels, S. et al.: Fast Bayesian Optimization of Machine Learning Hyperparameters on Large Datasets. ArXiv 2016, https://doi.org/10.48550/arXiv.1605.07079 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  15. [15] Koskela, A.; Kulkarni, T.: Practical Differentially Private Hyperparameter Tuning with Subsampling. ArXiv 2023, https://doi.org/10.48550/arXiv.2301.11989 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  16. [16] Jin, H.: Hyperparameter Importance for Machine Learning Algorithms. ArXiv 2022, https://doi.org/10.48550/arXiv.2201.05132 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  17. [17] Herzen, J.; Lässig, F.; Piazzetta, S. G. et al.: Darts: User-friendly forecasting for Time Series. Journal of Machine Learning Research 23 (2022) 124, pp. 1–6, jmlr.org/papers/v23/21–1177.html Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  18. [18] Chen, T.; Guestrin, C.: XGBoost: A Scalable Tree Boosting System. Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. 2016, pp. 785–794, https://doi.org/10.1145/2939672.2939785 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  19. [19] Bai, S.; Kolter, J. Z.; Koltun, V.: An Empirical Evaluation of Generic Convolutional and Recurrent Networks for Sequence Modeling. ArXiv 2018, https://doi.org/10.48550/arXiv.1803.01271 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  20. [20] Akiba, T.; Sano, S.; Yanase, T. et al.: Optuna: A Next-generation Hyperparameter Optimization Framework. ArXiv 2019, https://doi.org/10.48550/arXiv.1907.10902 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  21. [21] Victoria, A. H.; Maragatham, G.: Automatic tuning of hyperparameters using Bayesian optimization. Evolving Systems 12 (2021) 1, pp. 217–223, https://doi.org/10.1007/s12530–020–09345–2 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  22. [22] Vincent, A. M.; Jidesh, P.: An improved hyperparameter optimization framework for AutoML systems using evolutionary algorithms. Scientific Reports 13 (2023) 1, #4737, https://doi.org/10.1038/s41598-023-32027-3 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  23. [23] Abbas, A. T.; Ragab, A. E.; Benyahia, F. et al.: Taguchi Robust Design for Optimizing Surface Roughness of Turned AISI 1045 Steel Considering the Tool Nose Radius and Coolant as Noise Factors. Advances in Materials Science and Engineering 2018 (2018) 1, #2560253, https://doi.org/10.1155/2018/2560253 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  24. [24] Amir, B.; Gale, Y.; Sadot, A. et al.: Study on the effects of manufacturing parameters on the dynamic properties of AlSi10Mg under dynamic loads using Taguchi procedure. Materials & Design 223 (2022), #111125, https://doi.org/10.1016/j.matdes.2022.111125 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  25. [25] Khaw, J. F. C.; Lim, B. S.; Lim, L. E. N.: Optimal design of neural networks using the Taguchi method. Neurocomputing 7 (1995) 3, pp. 225–245, https://doi.org/10.1016/0925–2312(94)00013-I Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  26. [26] Groemping, U.: DoE.base: Full Factorials, Orthogonal Arrays and Base Utilities for DoE Packages. 2009, pp. 1.2–5, https://doi.org/10.32614/CRAN.package.DoE.base Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  27. [27] Mitra, A.: The Taguchi method. WIREs Computational Statistics 3 (2011) 5, pp. 472–480, https://doi.org/10.1002/wics.169 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  28. [28] Wu, L.: Mixed Effects Models for Complex Data. University of British Columbia 2009, www.stat.ubc.ca/~lang/Book/mybook.pdf Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  29. [29] Quirante, T.; Sebastian, P.; Ledoux, Y.: A trade-off function to tackle robust design problems in engineering. Journal of Engineering Design 24 (2013) 1, pp. 64–81, https://doi.org/10.1080/09544828.2012.691160 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51
  30. [30] Hisam, M. W.; Dar, A. A.; Elrasheed, M. O.; Khan, M. S.; Gera, R.; Azad, I.: The Versatility of the Taguchi Method: Optimizing Experiments Across Diverse Disciplines. Journal of Statistical Theory and Applications 23 (2024) 4, pp. 365–389, https://doi.org/10.1007/s44199-024-00093-9 Google Scholar öffnen DOI: 10.37544/1436-4980-2026-04-51

Zitation

Download RIS Download BibTex

Downloads

Download Kapitel (PDF)

Hinweis

Dieses Werk ist urheberrechtlich geschützt und darf nur nach den Vorgaben der Nutzungs- und Lizenzvereinbarung genutzt werden. Jede darüber hinaus gehende Nutzung des Werkes ist ausdrücklich untersagt und wird verfolgt.

Per E-Mail teilen

Medien