Catarina Dudas

Integration of data mining and multi-objective optimisation for decision support in production systems development

Multi-objective optimisation (MOO) is a powerful approach for generating a set of optimal trade-off (Pareto) design alternatives that the decision-maker can evaluate and then choose the most-suitable configuration, based on some high-level strategic information. Nevertheless, in practice, choosing among a large number of solutions on the Pareto front is often a daunting task, if proper analysis and visualisation techniques are not applied. Recent research advancements have shown the advantages of using data mining techniques to automate the post-optimality analysis of Pareto-optimal solutions for engineering design problems. Nonetheless, it is argued that the existing approaches are inadequate for generating high-quality results, when the set of the Pareto solutions is relatively small and the solutions close to the Pareto front have almost the same attributes as the Pareto-optimal solutions, of which both are commonly found in many real-world system problems. The aim of this paper is therefore to propose a distance-based data mining approach for the solution sets generated from simulation-based optimisation, in order to address these issues. Such an integrated data mining and MOO procedure is illustrated with the results of an industrial cost optimisation case study. Particular emphasis is paid to showing how the proposed procedure can be used to assist decision-makers in analysing and visualising the attributes of the design alternatives in different regions of the objective space, so that informed decisions can be made in production systems development.

Simulation-Based Innovization Using Data Mining for Production Systems Analysis

This chapter introduces a novel methodology for the analysis and optimization of production systems. The methodology is based on the innovization procedure, originally introduced for unveiling new and innovative design principles in engineering design problems. Although the innovization method is based on multi-objective optimization and post-optimality analyses of optimised solutions, it stretches the scope beyond an optimization task and attempts to discover new design/operational rules/principles relating to decision variables and objectives, so that a deeper understanding of the problem can be obtained. By integrating the concept of innovization with discrete-event simulation and data mining techniques, a new set of powerful tools can be developed for general systems analysis, particularly suitable for production systems. The uniqueness of the integrated approach proposed in this chapter lies on applying data mining to the data sets generated from simulation-based multi-objective optimization, in order to automatically or semi-automatically discover and interpret the hidden relationships and patterns for optimal production systems design/reconfiguration. After describing the simulation-based innovization using data mining procedure and its difference from conventional simulation analysis methods, results from an industrial case study carried out for the improvement of an assembly line in an automotive manufacturer will be presented.

Post-analysis of multi-objective optimization solutions using decision trees

Evolutionary algorithms are often applied to solve multi-objective optimization problems. Such algorithms effectively generate solutions of wide spread, and have good convergence properties. However, they do not provide any characteristics of the found optimal solutions, something which may be very valuable to decision makers. By performing a post-analysis of the solution set from multi-objective optimization, relationships between the input space and the objective space can be identified. In this study, decision trees are used for this purpose. It is demonstrated that they may effectively capture important characteristics of the solution sets produced by multi-objective optimization methods. It is furthermore shown that the discovered relationships may be used for improving the search for additional solutions. Two multi-objective problems are considered in this paper; a well-studied benchmark function problem with on a beforehand known optimal Pareto front, which is used for verification purposes, and a multi-objective optimization problem of a real-world production system. The results show that useful relationships may be identified by employing decision tree analysis of the solution sets from multi-objective optimizations.

Interleaving Innovization with Evolutionary Multi-Objective Optimization in Production System Simulation for Faster Convergence

This paper introduces a novel methodology for the optimization, analysis and decision support in production systems engineering. The methodology is based on the innovization procedure, originally introduced to unveil new and innovative design principles in engineering design problems. The innovization procedure stretches beyond an optimization task and attempts to discover new design/operational rules/principles relating to decision variables and objectives, so that a deeper understanding of the underlying problem can be obtained. By integrating the concept of innovization with simulation and data mining techniques, a new set of powerful tools can be developed for general systems analysis. The uniqueness of the approach introduced in this paper lies in that decision rules extracted from the multi-objective optimization using data mining are used to modify the original optimization. Hence, faster convergence to the desired solution of the decision-maker can be achieved. In other words, faster convergence and deeper knowledge of the relationships between the key decision variables and objectives can be obtained by interleaving the multi-objective optimization and data mining process. In this paper, such an interleaved approach is illustrated through a set of experiments carried out on a simulation model developed for a real-world production system analysis problem.

Using uncertain chemical and thermal data to predict product quality in a casting process

Process and casting data from different sources have been collected and merged for the purpose of predicting, and determining what factors affect, the quality of cast products in a foundry. One problem is that the measurements cannot be directly aligned, since they are collected at different points in time, and instead they have to be approximated for specific time points, hence introducing uncertainty. An approach for addressing this problem is investigated, where uncertain numeric feature values are represented by intervals and random forests are extended to handle such intervals. A preliminary experiment shows that the suggested way of forming the intervals, together with the extension of random forests, results in higher predictive performance compared to using single (expected) values for the uncertain features together with standard random forests.