MCDM'07 - paper no. 11

 

Back to MCDM'07 contents
 

Representing partial information on preferences with the help of linear transformation of objective space

Dmitry Podkopaev

Abstract:

We consider the case in which information about preferences in a multi-objective optimization problem is represented in the form of upper bounds on trade-off coefficients, defined for any pair of objective functions. We introduce some requirements for this preference system and derive a system of constraints which should be put on the upper bounds. We use a matrix whose elements are numbers inverse to upper bounds on trade-offs as the linear operator for transforming the objective space. The presented constraint system allows for proving that if a given evaluation is weakly Pareto optimal in the transformed space, then in the initial space it is Pareto optimal and satisfies the corresponding bounds on trade-off coefficients.

Keywords:

multiple objective optimization, incomplete preference information, bounds on trade-off coefficients, linear transformation of objective space

Reference index:

Dmitry Podkopaev, (2008), Representing partial information on preferences with the help of linear transformation of objective space, Multiple Criteria Decision Making (3), pp. 175-194

Full text:

download

Scopus citations in 1 paper(s):
  1. Podkopaev, D., & Miettinen, K. (2011). Handling preferences in the "pre-conflicting" phase of decision making processes under multiple criteria doi:10.1007/978-3-642-24873-3_18