Comparing graphic and symbolic classification in interactive multiobjective optimization

Abstract

In interactive multiobjective optimization systems, the classification of objective functions is a convenient way to direct the solution process in order to search for new, more satisfactory, solutions in the set of Pareto optimal solutions. Classification means that the decision maker assigns the objective functions into classes depending on what kind of changes in their values (in relation to the current values) are desirable.Here we study the role of user interfaces in implementing classification in multiobjective optimization software and how classification should be realized. In this way, we want to pay attention to the usability of multiobjective optimization software. Typically, this topic has not been of interest in the multiobjective optimization literature. However, usability aspect is important because in interactive classification-based multiobjective optimization methods, the classification is the core of the solution process. We can say that the more convenient the classification is, the more efficient the system or the method is and the better it supports the work of the decision maker.We report experiments with two classification options, graphic and symbolic ones, which are used in connection with an interactive multiobjective optimization system WWW-NIMBUS. The ideas and conclusions given are applicable for other interactive classification-based method, as well.

Introduction to Multiobjective Optimization: Interactive Approaches

“…The most important aspect of WWW-NIMBUS R is that it is easily accessible and available to any academic Internet user at http://nimbus.it.jyu.fi/. For a discussion on how to design user interfaces for a software implementing a classification-based interactive method, see (Miettinen and Kaario, 2003). (When the first version of WWW-NIMBUS R was implemented in 1995 it was a pioneering interactive optimization system on the Internet.)…”

Interactive Optimization With Parallel Coordinates: Exploring Multidimensional Spaces for Decision Support

“…Branke et al (2008, p. 52) explicitly mention the importance of user-friendliness in IO as a topic for future research. This is also true for the representation of interaction with the problem, e.g., how the DM inputs their preferences (Miettinen and Kaario, 2003), and the representation of the preferences themselves (Branke et al, 2008, p. 201). Liu et al (2018) note that despite interactive optimization being "essentially a visual analytics task, " literature is rather silent on the specifics of visuals and interaction approaches, focusing rather on optimization procedures and preference models.…”

Implementation aspects of interactive multiobjective optimization for modeling environments: the case of GAMS-NIMBUS

“…Thus, in this research we focus on the algorithm side and do not consider user interface implementation (for such studies see, e.g., [32,48]). Besides, user interfaces may need application-specic elements but we want to retain on a more general algorithm level without going into application-specic details.…”