Bandaru, S., & Ng, A. H. C. (2019). Trend mining: A visualization technique to discover variable trends in the objective space. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 11411 LNCS, pp. 605–617). Springer Verlag.
Abstract
Practical multi-objective optimization problems often involve several decision variables that influence the objective space in different ways. All variables may not be equally important in determining the trade-offs of the problem. Decision makers, who are usually only concerned with the objective space, have a hard time identifying such important variables and understanding how the variables impact their decisions and vice versa. Several graphical methods exist in the MCDM literature that can aid decision makers in visualizing and navigating high-dimensional objective spaces. However, visualization methods that can specifically reveal the relationship between decision and objective space have not been developed so far. We address this issue through a novel visualization technique called trend mining that enables a decision maker to quickly comprehend the effect of variables on the structure of the objective space and easily discover interesting variable trends. The method uses moving averages with different windows to calculate an interestingness score for each variable along predefined reference directions. These scores are presented to the user in the form of an interactive heatmap. We demonstrate the working of the method and its usefulness through a benchmark and two engineering problems. Access publication.