Unsupervised Learning and Simulation for Complexity Management in Business Operations
A key resource in data analytics projects is the data to be analyzed. What can be done in the middle of a project if this data is not available as planned? This chapter explores a potential solution based on a use case from the manufacturing industry where the drivers of production complexity (and thus costs) were supposed to be determined by analyzing raw data from the shop floor, with the goal of subsequently recommending measures to simplify production processes and reduce complexity costs.
The unavailability of the data—often a major threat to the anticipated outcome of a project—has been alleviated in this case study by means of simulation and unsupervised machine learning: a physical model of the shop floor produced the necessary lower-level records from high-level descriptions of the facility. Then, neural autoencoders learned a measure of complexity regardless of any human-contributed labels.
In contrast to conventional complexity measures based on business analysis done by consultants, our data-driven methodology measures production complexity in a fully automated way while maintaining a high correlation to the human-devised measures.
Unable to display preview. Download preview PDF.
The authors are grateful for the support by CTI grant 18993.1 PFES-ES, and for the participation of our colleagues from the ZHAW Datalab in the conducted survey.
- Abbeel, P. (2017). Pieter Abbeel: Deep learning-to-learn robotic control [video-file]. Retrieved from https://youtu.be/TERCdog1ddE
- Budde, L., Faix, A., & Friedli, T. (2015). From functional to cross-functional management of product portfolio complexity. In Presented at the POMS 26th Annual Conference, Washington, DC.Google Scholar
- Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. Cambridge, MA: MIT Press. Retrieved December 22, 2017, from http://www.deeplearningbook.org
- Kawaguchi, K., Kaelbling, L. P., & Bengio, Y. (2017). Generalization in deep learning. CoRR, 1710, 05468. Retrieved December 22, 2017, from https://arxiv.org/abs/1710.05468
- Lichtensteiger, L. & Pfeifer, R. (2002). An optimal sensor morphology improves adaptability of neural network controllers. In J. R. Dorronsoro (Ed.), Proceedings of the International Conference on Artificial Neural Networks (ICANN 2002), Lecture Notes in Computer Science LNCS 2415 (pp. 850–855).Google Scholar
- Meierhofer, J. & Meier, K., (2017). From data science to value creation. In St. Za, M. Drăgoicea, & M. Cavallari (Eds.) Exploring Services Science, 8th International Conference, IESS 2017, Rome, Italy, May 24–26, 2017, Proceedings (pp. 173–181). Cham: Springer.Google Scholar
- Pimentel, D., Nowak, R., & Balzano, L. (2014, June). On the sample complexity of subspace clustering with missing data. In 2014 IEEE Workshop on Statistical Signal Processing (SSP) (pp. 280–283). IEEE.Google Scholar
- Ramdas, K., & Sawhney, M. S. (2001, January 1). A cross-functional approach to evaluating multiple line extensions for assembled products [research-article]. Retrieved January 15, 2014, from http://pubsonline.informs.org/doi/abs/10.1287/mnsc.126.96.36.19967
- Schmidhuber, J. (2008). Driven by compression progress: A simple principle explains essential aspects of subjective beauty, novelty, surprise, interestingness, attention, curiosity, creativity, art, science, music, jokes. In Workshop on anticipatory behavior in adaptive learning systems (pp. 48–76). Heidelberg: Springer.Google Scholar
- Stadelmann, T., Tolkachev, V., Sick, B., Stampfli, J., & Dürr, O. (2018). Beyond ImageNet - deep learning in industrial practice. In M. Braschler, T. Stadelmann, & K. Stockinger (Eds.), Applied data science: Lessons learned for the data-driven business. Heidelberg: Springer.Google Scholar
- Swiss Alliance for Data-Intensive Services. (2017). Digitization & innovation through cooperation. Glimpses from the digitization & innovation workshop at “Konferenz Digitale Schweiz”. Retrieved April 26, 2018, from https://data-service-alliance.ch/blog/blog/digitization-innovation-through-cooperation-glimpses-from-the-digitization-innovation-workshop
- Zhu, X., Gibson, B. R., & Rogers, T. T. (2009). Human rademacher complexity. In Advances in neural information processing systems (pp. 2322–2330). Cambridge, MA: MIT Press.Google Scholar