A synergy of multi-objective optimization and data mining for the analysis of a flexible flow shop

Authors:
Catarina Dudas;Marcus Frantzén;Amos H. C. Ng
Affiliations:
Virtual Systems Research Centre, University of Skövde, Skövde 541 28, Sweden;Virtual Systems Research Centre, University of Skövde, Skövde 541 28, Sweden;Virtual Systems Research Centre, University of Skövde, Skövde 541 28, Sweden
Venue:
Robotics and Computer-Integrated Manufacturing
Year:
2011

Citing 5
Cited 0

C4.5: programs for machine learning

C4.5: programs for machine learning
An Empirical Comparison of Voting Classification Algorithms: Bagging, Boosting, and Variants

Machine Learning
Induction of Decision Trees

Machine Learning
Innovization: innovating design principles through optimization

Proceedings of the 8th annual conference on Genetic and evolutionary computation
A fast and elitist multiobjective genetic algorithm: NSGA-II

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

A method for analyzing production systems by applying multi-objective optimization and data mining techniques on discrete-event simulation models, the so-called Simulation-based Innovization (SBI) is presented in this paper. The aim of the SBI analysis is to reveal insight on the parameters that affect the performance measures as well as to gain deeper understanding of the problem, through post-optimality analysis of the solutions acquired from multi-objective optimization. This paper provides empirical results from an industrial case study, carried out on an automotive machining line, in order to explain the SBI procedure. The SBI method has been found to be particularly suitable in this case study as the three objectives under study, namely total tardiness, makespan and average work-in-process, are in conflict with each other. Depending on the system load of the line, different decision variables have been found to be influencing. How the SBI method is used to find important patterns in the explored solution set and how it can be valuable to support decision making in order to improve the scheduling under different system loadings in the machining line are addressed.