╨╧рб▒с>■  &(■   %                                                                                                                                                                                                                                                                                                                                                                                                                                                ье┴q`Ё┐gbjbjqPqP8::g       дxxxxxxxМ  МЙЖ888888N Z      $ hw ^.xb88bb.xx88C┤┤┤bx8x8┤b┤┤xx┤8, |KFк│╥x"┤№ Y0Й┤╒ Ъ╒ ┤┤╒ x─8bb┤bbbbb..к bbbЙbbbbМММD╨DМММ╨МММxxxxxx     Tэtulo: Integrating Collective Intelligence into Multi-Objective Optimization Evolutionary Algorithms: Interactive Preferences and Reference Points for a Facility Location Problem Abstract: Some organizations have to assign and manage facilities in an optimized way. Those activities involve many stakeholders with multiple conflicting objectives. Multi-objective optimization evolutionary algorithms have been successfully applied to several complex synthetic and real-world multi-objective problems (MOPs). Although these algorithms have proved themselves as a valid approach to the MOP, there is still need for improvements on the performance of the search process. This work introduces a novel approach meant for bringing collective intelligence methods into the optimization process carried out by evolutionary multi-objective optimization algorithms. In particular, it describes the extension of some well-known algorithms (Non-dominated Sorting Genetic Algorithm-II, S-metric Selection Evolutionary Multi-objective Algorithm, Strength Pareto Evolutionary Algorithm 2) to include collective online preferences into the optimization process. In these new methods Чcalled CI-NSGA-II, CI-SMS-EMOA and CI-SPEA2Ч, groups of decision makers can highlight the regions of the Pareto frontier that are more relevant to them as to focus the search process mainly on those areas. Additionally, the integration of interactivity and cooperation into the evolutionary algorithms refines usersТ preferences and improves the reference points throughout the evolutionary progress. Rather than a unique or small group of decision makers with unilateral preferences, the application of dynamic group preferences aggregates consistent collective reference points and creative solutions to enhance multi-objective results. In order to analyze the results, three new performance indicators based on preferences are introduced to evaluate the quality of approximations set. As part of this work, the algorithmsТ performances are tested when faced with some synthetic problems as well as a real-world case of facility location. The experiments demonstrate the advantages of a collective intelligence operator integrated into the multi-objective evolutionary algorithm. Keywords: collective intelligence; preferences; reference points; evolutionary multi- objective optimization algorithms; facility location problem.  ╡╕┬├╤┘efgэ▄├оШо├}├h[hўБh#ы^JmH sH )hи h#ыB*OJQJ^JmH phsH 4hи h#ы5БB*CJOJQJ^JaJmH phsH +hи h#ы5БCJOJQJ^JaJmH sH (hи h#ыCJOJQJ^JaJmH sH 1hи h#ыB*CJOJQJ^JaJmH phsH  hи h#ыCJOJQJ^JaJ#hи h#ы5БCJOJQJ^JaJ  ╢╖╕┬├╤fg¤¤█¤¤¤¤╚╕¤dhдЁ1$7$8$H$gd7F$dhдЁ1$7$8$H$a$gd7F"$ ╞2Ф(╝ Pфx а4 ╚#\'Ё*Д.2м5@9a$gd7F g¤61РhP:pтA░|. ░╚A!░е"░е#РЙ$РЙ%░░─░─ Р─ЖЬ 666666666666666666666666666666666666666666 6666666666 666666666666 666666666666666666666666666666666666666666666666666666666666666666D@ё D ╤+ёNormalCJ^J_HaJmHsHtH >AЄ б> 0Fonte parсg. padrуoTiє │T 0 Tabela normalЎ4╓ l4╓aЎ ,kЇ ┴, 0 Sem lista МeЄМ 7F0Prщ-formataчуo HTML7 ╞2Ф(╝ Pфx а4 ╚#\'Ё*Д.2м5@9CJOJQJ^JaJ@■в@ 7F0 Char CharCJOJQJ^JaJg      ╢╖╕┬├╤f i Ш0ААШ@0ААШ@0ААШ@0ААШ@0ААШ@0ААШ@0ААШ@0ААШ0ААШ0ААi K╚00KИ00L3g g g   _GoBacki i √ i -8i 33i i `х7Fи тAўБrПШl▓#ы╤+ёi  @╤┘(&╤╤g `@  Unknown            GР :рAx└  Times New Roman5РАSymbol3&Р  :рCx└  ArialGРАMS ??MS Mincho7Р р @ЯCambriaO1Р CourierCourier New"AИЁ╨йu║SЗ_cTg xяxябЁеЙ┤┤ББ24c c @Ё№ (Ё $P                     7F2  Tэtulo:Daniel CinalliHelio■ рЕЯЄ∙OhлС+'│┘0tРШи┤╠╪фЇ  0 < HT\dlфTэtulo:Daniel CinalliNormalHelio9Microsoft Office Word@q@▐╚f╘г╥@zU+к│╥xя■ ╒═╒Ь.УЧ+,∙о0° hpДМФЬ дм┤╝ ─ ╪ф QuatroSemc ц Tэtulo: Tэtulo ■   ■   ■    !"#$■   ¤   '■   ■   ■                                                                                                                                                                                                                                                                                                                                                           Root Entry         └FАoQFк│╥)А1Table         WordDocument        8SummaryInformation(    DocumentSummaryInformation8            CompObj            u                        ■                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                           ■       └F#Documento do Microsoft Office Word MSWordDocWord.Document.8Ї9▓q