Infusing business optimization processes with machine learning and expert knowledge

Summary

Leaving in the 4th industrial revolution, the digitization in manufacturing in conjunction with the rapid increase of AI applications have raised new opportunities. In the context of this work, the infusion of machine learning on a real word manufacturing optimization knapsack-like problem will be researched and addressed by focusing on interpretability and generalization of the models. The problem will be described, modelized and formulated mathematically, the data collection and dataset construction will be presented, and machine learning techniques will be used and compared based on which achieves to extract the most information out of the data. Also, the explanation of the trained models is discussed, and the most efficient way of data constructions is checked in terms of generating a model that generalizes better.