Research Article
Comparing Solutions under Uncertainty in
Multiobjective Optimization
Miha Mlakar, Tea Tušar, and Bogdan FilipiI
Department of Intelligent Systems, Joˇ zef Stefan Institute and Joˇ zef Stefan International Postgraduate School,
Jamova cesta 39, 1000 Ljubljana, Slovenia
Correspondence should be addressed to Miha Mlakar; miha.mlakar@ijs.si
Received 19 December 2013; Revised 14 April 2014; Accepted 17 April 2014; Published 18 May 2014
Academic Editor: Zhan Shu
Copyright © 2014 Miha Mlakar et al. Tis is an open access article distributed under the Creative Commons Attribution License,
which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Due to various reasons the solutions in real-world optimization problems cannot always be exactly evaluated but are sometimes
represented with approximated values and confdence intervals. In order to address this issue, the comparison of solutions has to be
done diferently than for exactly evaluated solutions. In this paper, we defne new relations under uncertainty between solutions in
multiobjective optimization that are represented with approximated values and confdence intervals. Te new relations extend the
Pareto dominance relations, can handle constraints, and can be used to compare solutions, both with and without the confdence
interval. We also show that by including confdence intervals into the comparisons, the possibility of incorrect comparisons, due
to inaccurate approximations, is reduced. Without considering confdence intervals, the comparison of inaccurately approximated
solutions can result in the promising solutions being rejected and the worse ones preserved. Te efect of new relations in the
comparison of solutions in a multiobjective optimization algorithm is also demonstrated.
1. Introduction
Multiobjective optimization is the process of simultaneously
optimizing two or more conficting objectives. Problems
with multiple objectives can be found in various felds,
from product design and process optimization to fnancial
applications. Teir specifcity is that the result is not just
one solution, but a set of solutions representing trade-ofs
between objectives.
Multiobjective evolutionary algorithms (MOEAs) are
known for efciently solving these kind of problems [1].
However, MOEAs can also be used for solving optimization
problems with uncertain objective values. Te reason for
uncertainty can be noise, robustness, ftness approximations,
or time-varying ftness functions. When solving uncertain
optimization problems, it is better if the algorithm takes
uncertainty into account.
Uncertain solutions can be represented with approxi-
mated values and variances of these approximations. From
the variance, the confdence interval of the approximation
can be calculated. Tis interval indicates the region in which
the exactly evaluated solution is most likely to appear. Te
confdence interval width indicates the certainty of the
approximation. If the confdence interval is narrow, we can
be more certain about the approximation and vice versa.
Since the confdence intervals ofer additional information on
the approximations, they can be efectively used to compare
solutions and an algorithm using confdence intervals can
perform better by exploiting this additional information
[2]. During optimization that does not consider confdence
intervals, an approximated solution may be incorrectly iden-
tifed as the better of the two compared solutions. Ofen
the solution that is incorrectly determined as worse is then
discarded. Similarly, a promising solution can get discarded
if a worse solution is incorrectly determined as being better.
In both cases good solutions are lost due to the comparison
of solutions which only considers approximated values.
To prevent these unwanted efects, we propose new
relations for comparing solutions under uncertainty, where,
in addition to the approximated values of a solution, their
confdence intervals are considered. Tese relations cover
all possible combinations that can occur when comparing
Hindawi Publishing Corporation
Mathematical Problems in Engineering
Volume 2014, Article ID 817964, 10 pages
http://dx.doi.org/10.1155/2014/817964