Selection in Evolutionary Multiobjective Optimization

Data
11/09/20152015-09-10 21:00:00 2015-09-10 21:00:00 Selection in Evolutionary Multiobjective Optimization The potential of evolutionary algorithms in multiobjective optimization was identified early in their history. That potential has been realized over the years with the development of increasingly elaborate Evolutionary Multiobjective Optimization (EMO) algorithms that have found many important applications in the real world, and have contributed significantly to the growth in popularity of multiobjective optimization in general.  Although EMO has traditionally emphasized the approximation of the whole Pareto-optimal front in an a posteriori articulation of preferences setting, preference-driven EMO algorithms capable of handling interaction with a Decision Maker (DM) were proposed early in EMO development. While the identification of a most preferred solution is usually seen as the ultimate goal in practice, recognizing that the search for diverse sets of alternative solutions to be presented to the DM (whether in a progressive or an a posteriori articulation of preferences scenario) implies some sort of set-oriented preference for diversity has been a turning point in EMO algorithm development. State-of-the-art algorithms such as IBEA, SMS-EMOA, MO-CMA-ES and HyPE, for example, implement multiobjective selection based on a notion of set quality that is then used to infer the quality of the individuals in the population and to introduce bias towards the better ones at the parental and/or environmental selection stages. However, how to combine such set-oriented preferences with the more traditional search for a single most-preferred solution remains largely an open question.  This talk focuses on the problem of selecting a diverse subset of non-dominated solutions from a larger set of candidate solutions according to DM preference information. The expression of set-oriented preferences by the DM, their incorporation in EMO algorithms, and computational aspects of the resulting subset selection problems are considered. Existing quality-indicator and decomposition approaches are reviewed and discussed, and an alternative perspective is introduced where set quality is not specified as such by the DM, but is inferred from the uncertainty associated with DM solution-oriented preferences instead. Recent results obtained by instantiating this idea in the form of a portfolio optimization problem are presented and discussed, and opportunities for further work are outlined at the end.  IC 3,5 - Sala 351 IC 3,5 - Sala 351 IC 3,5 - Sala 351 America/Sao_Paulo public
Horário
14:00 h
Local
IC 3,5 - Sala 351
Palestrante
Prof. Carlos M. Fonseca (CISUC, Department of Informatics Engineering, University of Coimbra, Portugal)
Descrição

The potential of evolutionary algorithms in multiobjective optimization was identified early in their history. That potential has been realized over the years with the development of increasingly elaborate Evolutionary Multiobjective Optimization (EMO) algorithms that have found many important applications in the real world, and have contributed significantly to the growth in popularity of multiobjective optimization in general. 

Although EMO has traditionally emphasized the approximation of the whole Pareto-optimal front in an a posteriori articulation of preferences setting, preference-driven EMO algorithms capable of handling interaction with a Decision Maker (DM) were proposed early in EMO development. While the identification of a most preferred solution is usually seen as the ultimate goal in practice, recognizing that the search for diverse sets of alternative solutions to be presented to the DM (whether in a progressive or an a posteriori articulation of preferences scenario) implies some sort of set-oriented preference for diversity has been a turning point in EMO algorithm development. State-of-the-art algorithms such as IBEA, SMS-EMOA, MO-CMA-ES and HyPE, for example, implement multiobjective selection based on a notion of set quality that is then used to infer the quality of the individuals in the population and to introduce bias towards the better ones at the parental and/or environmental selection stages. However, how to combine such set-oriented preferences with the more traditional search for a single most-preferred solution remains largely an open question. 

This talk focuses on the problem of selecting a diverse subset of non-dominated solutions from a larger set of candidate solutions according to DM preference information. The expression of set-oriented preferences by the DM, their incorporation in EMO algorithms, and computational aspects of the resulting subset selection problems are considered. Existing quality-indicator and decomposition approaches are reviewed and discussed, and an alternative perspective is introduced where set quality is not specified as such by the DM, but is inferred from the uncertainty associated with DM solution-oriented preferences instead. Recent results obtained by instantiating this idea in the form of a portfolio optimization problem are presented and discussed, and opportunities for further work are outlined at the end. 

Sobre o Palestrante

Carlos M. Fonseca is an Associate Professor at the Department of Informatics Engineering of the University of Coimbra, Portugal, and the Head of the Evolutionary and Complex Systems (ECOS) group of the Centre for Informatics and Systems of the University of Coimbra (CISUC). He graduated in Electronic and Telecommunications Engineering from the University of Aveiro, Portugal, in 1991, and obtained a Ph.D. in Automatic Control and Systems Engineering from the University of Sheffield, U.K., in 1996. He was a Research Associate with the Department of Automatic Control and Systems Engineering of the University of Sheffield from 1994 until 1997, and a Lecturer at the Department of Electronic Engineering and Informatics, Faculty of Science and Technology, University of Algarve, Faro, Portugal, from 1997 until October 2010, when he joined the University of Coimbra. 

His research has been devoted mainly to evolutionary computation and multi-objective optimization. In the 1990's, he proposed MOGA, a "first-generation" multi-objective evolutionary algorithm with support for progressive preference articulation through goals and priorities, and began the development of the attainment-function approach to the experimental evaluation of stochastic multi-objective optimization algorithms. Since then, he has focused on the study of evolutionary algorithm dynamics, including representation and convergence aspects; further development of statistical methodologies for the experimental evaluation of multi-objective optimization algorithms, and of efficient algorithms to support them; and the development of new, computationally efficient, approaches to preference articulation in evolutionary multi-objective optimization. He has also contributed to several applications in the engineering and management domains. 

He was a General co-Chair of the International Conference on Evolutionary Multi-Criterion Optimization (EMO) in 2003, 2009 and 2013, and a Technical co-Chair of the IEEE Congress on Evolutionary Computation (CEC) in 2000 and 2005. He is a member of the Evolutionary Multi-Criterion Optimization Steering Committee and a member of the International Society on Multiple Criteria Decision Making (MCDM), the Portuguese Operations Research Association (APDIO), and the Portuguese Association of Automatic Control.

Informações Adicionais

Responsável: Profa. Ariadne M. B. R. Carvalho
Email: ariadne@ic.unicamp.br
Fone: (19) 3521-5864
Instituto de Computação, Unicamp