www.wikidata.uk-ua.nina.az
Opukla optimizaciya ce pidrozdil matematichnoyi optimizaciyi kotrij vivchaye problemu minimizaciyi opuklih funkcij nad opuklimi mnozhinami Bagato klasiv zadach z opukloyu optimizaciyeyu dopuskayut polinomialni algoritmi 1 todi yak matematichna optimizaciya v cilomu NP vazhka 2 3 4 Opukla optimizaciya maye zastosuvannya v shirokomu spektri disciplin takih yak avtomatichni sistemi upravlinnya ocinka ta obrobka signaliv komunikaciyi ta merezhi proektuvannya elektronnih shem 5 analiz ta modelyuvannya danih finansi statistika optimalnij eksperimentalnij dizajn 6 ta strukturna optimizaciya de koncepciya nablizhennya viyavilas efektivnoyu 5 7 Z nedavnimi dosyagnennyami v galuzi obchislyuvalnih ta optimizacijnih algoritmiv opukle programuvannya majzhe nastilki zh proste yak i linijne programuvannya 5 Zmist 1 Viznachennya 1 1 Standartna forma 2 Vlastivosti 3 Prikladi 4 Mnozhniki Lagranzha 5 Algoritmi 6 Rozshirennya 7 Div takozh 8 Primitki 9 Spisok literaturi 10 PosilannyaViznachennya RedaguvatiProblema optimizaciyi opuklosti ce problema optimizaciyi v yakij cilova funkciya ye opukloyu funkciyeyu a dopustimoyu mnozhinoyu ye opukla mnozhina Funkciya f displaystyle f nbsp vidobrazhennya deyakoyi pidmnozhini R n displaystyle mathbb R n nbsp v R displaystyle mathbb R cup pm infty nbsp opukla yaksho yiyi domen opuklij i dlya vsih 8 0 1 displaystyle theta in 0 1 nbsp i takozh dlya vsih x y displaystyle x y nbsp u svoyemu domeni vikonuyetsya taka umova f 8 x 1 8 y 8 f x 1 8 f y displaystyle f theta x 1 theta y leq theta f x 1 theta f y nbsp Mnozhina S opukla yaksho dlya vsih chleniv x y S displaystyle x y in S nbsp i dlya vsih 8 0 1 displaystyle theta in 0 1 nbsp u nas ye sho 8 x 1 8 y S displaystyle theta x 1 theta y in S nbsp Konkretno problema opukloyi optimizaciyi ce problema poshuku x C displaystyle mathbf x ast in C nbsp mayuchi inf f x x C displaystyle inf f mathbf x mathbf x in C nbsp de ob yektivna funkciya f displaystyle f nbsp ye opukloyu yak i dopustima mnozhina C displaystyle C nbsp 8 9 Yaksho taka tochka isnuye vona nazivayetsya optimalnoyu tochkoyu mnozhina vsih optimalnih tochok nazivayetsya optimalnoyu mnozhinoyu Yaksho f displaystyle f nbsp ye neobmezhenoyu vnizu nad C displaystyle C nbsp abo minimum ne dosyagnuto todi yak kazhut problema optimizaciyi ye neobmezhenoyu Inakshe yaksho C displaystyle C nbsp ye porozhnoyu mnozhinoyu todi problema yak kazhut nevirishuvana 5 Standartna forma Redaguvati Kazhut sho problema opukloyi optimizaciyi ye v standartnij formi yaksho vona zapisana yak minimize x f x s u b j e c t t o g i x 0 i 1 m h i x 0 i 1 p displaystyle begin aligned amp underset mathbf x operatorname minimize amp amp f mathbf x amp operatorname subject to amp amp g i mathbf x leq 0 quad i 1 dots m amp amp amp h i mathbf x 0 quad i 1 dots p end aligned nbsp de x R n displaystyle x in mathbb R n nbsp zminna optimizaciyi funkciyi f g 1 g m displaystyle f g 1 ldots g m nbsp ye opuklimi i funkciyi h 1 h p displaystyle h 1 ldots h p nbsp ye afinnimi 5 U comu poznachenni funkciya f displaystyle f nbsp ce cilova funkciya zadachi i funkciyi g i displaystyle g i nbsp i h i displaystyle h i nbsp nazivayutsya funkciyami obmezhennya Mozhlivim naborom zadachi optimizaciyi ye mnozhina sho skladayetsya z usih tochok x R n displaystyle x in mathbb R n nbsp zadovolnyayuchi g 1 x 0 g m x 0 displaystyle g 1 x leq 0 ldots g m x leq 0 nbsp i h 1 x 0 h p x 0 displaystyle h 1 x 0 ldots h p x 0 nbsp Cya mnozhina ye opukloyu oskilki pidmnozhini opuklih funkcij opukli afinni mnozhini opukli a peretin opuklih mnozhin opuklij 5 Bagato problem optimizaciyi mozhut buti sformulovani v cij standartnij formi Napriklad problema maksimizaciyi uvignutoyi funkciyi f displaystyle f nbsp mozhe buti pereformulovano yak problema minimizaciyi opukloyi funkciyi f displaystyle f nbsp taka problema maksimizaciyi uvignutoyi funkciyi nad opukloyu mnozhinoyu chasto nazivayetsya problemoyu optimizaciyi opukloyi formi Vlastivosti RedaguvatiNastupni korisni vlastivosti zadach opukloyi optimizaciyi 5 10 kozhen lokalnij minimum ce globalnij minimum optimalna mnozhina opukla yaksho cilova funkciya strogo opukla to zadacha maye shonajmenshe odnu optimalnu tochku Ci rezultati vikoristovuyutsya teoriyeyu opukloyi minimizaciyi razom z geometrichnimi ponyattyami funkcionalnogo analizu v prostorah Gilberta takimi yak teorema proyekciyi Gilberta teorema rozdilennya giperplan ta lemma Farkasa Prikladi RedaguvatiPerelicheni klasi zadach ce vse zadachi opukloyi optimizaciyi abo yih mozhna zvesti do zadachi opukloyi optimizaciyi za dopomogoyu prostih peretvoren 5 11 nbsp Iyerarhiya zadach opukloyi optimizaciyi LP linijna programa QP kvadratichna programa programa konusiv SOCP drugogo poryadku SDP napivviznachena programa CP programa konusa Najmenshi kvadrati Linijne programuvannya Opukla kvadratichna minimizaciya z linijnimi obmezhennyami Kvadratna minimizaciya z opuklimi kvadratichnimi obmezhennyami Konichna optimizaciya Geometrichne programuvannya Programuvannya konusa drugogo poryadku Napivskinchenne programuvannya Maksimizaciya entropiyi z vidpovidnimi obmezhennyamiMnozhniki Lagranzha RedaguvatiRozglyanemo problemu minimizaciyi opukloyi formi zadanu v standartnij formi funkciyeyu vitrat f x displaystyle f x nbsp ta obmezhennyam nerivnosti g i x 0 displaystyle g i x leq 0 nbsp dlya 1 i m displaystyle 1 leq i leq m nbsp Domen X displaystyle mathcal X nbsp ye X x X g 1 x g m x 0 displaystyle mathcal X left x in X vert g 1 x ldots g m x leq 0 right nbsp Funkciyeyu Lagranzha dlya zadachi ye L x l 0 l 1 l m l 0 f x l 1 g 1 x l m g m x displaystyle L x lambda 0 lambda 1 ldots lambda m lambda 0 f x lambda 1 g 1 x cdots lambda m g m x nbsp Dlya kozhnoyi tochki x displaystyle x nbsp v X displaystyle X nbsp sho minimizuye f displaystyle f nbsp nad X displaystyle X nbsp isnuyut realni chisla l 0 l 1 l m displaystyle lambda 0 lambda 1 ldots lambda m nbsp kotri nazivayutsya mnozhnikami Lagranzha yaki odnochasno zadovolnyayut ci umovi x displaystyle x nbsp minimizuye L y l 0 l 1 l m displaystyle L y lambda 0 lambda 1 ldots lambda m nbsp nad usim y X displaystyle y in X nbsp l 0 l 1 l m 0 displaystyle lambda 0 lambda 1 ldots lambda m geq 0 nbsp prinajmni z odnim l k gt 0 displaystyle lambda k gt 0 nbsp l 1 g 1 x l m g m x 0 displaystyle lambda 1 g 1 x cdots lambda m g m x 0 nbsp dodatkova mlyavist Yaksho isnuye strogo dopustima tochka tobto tochka z displaystyle z nbsp kotra zadovolnyaye g 1 z g m z lt 0 displaystyle g 1 z ldots g m z lt 0 nbsp todi tverdzhennya vishe mozhe vimagati l 0 1 displaystyle lambda 0 1 nbsp I navpaki yaksho yakis x displaystyle x nbsp v X displaystyle X nbsp zadovolnyayut 1 3 dlya skalyariv l 0 l m displaystyle lambda 0 ldots lambda m nbsp z l 0 1 displaystyle lambda 0 1 nbsp to x displaystyle x nbsp minimizuye f displaystyle f nbsp nad X displaystyle X nbsp Algoritmi RedaguvatiZadachi opukloyi optimizaciyi mozhut buti rozv yazani takimi suchasnimi metodami 12 Metodi rozsharuvannya Vulf Lemarel Kivil ta Metodi subgradiyentnoyi proyekciyi Polyak Metodi vnutrishnih tochok 1 v yakih vikoristovuyutsya samokoreguyuchi bar yerni funkciyi 13 ta samoregulyarni bar yerni funkciyi 14 Rizhuchi ploshinni metodi Elipsoyidnij metod Subgradiyentnij metod Podvijni subgradiyenti ta metod drejfu plyus shtrafSubgradiyentni metodi mozhut buti realizovani prosto i tomu shiroko zastosovuyutsya 15 Podvijni subgradiyentni metodi ce subgradiyentni metodi zastosovani do podvijnoyi zadachi Metod drejfuvannya plyus shtrafu shozhij z metodom podvijnogo subgradiyenta Rozshirennya RedaguvatiRozshirennya opukloyi optimizaciyi vklyuchayut optimizaciyu funkcij dvoopukloyi psevdo opukloyi ta kvazi opukloyi Rozshirennya teoriyi opuklogo analizu ta iterativnih metodiv priblizno rozv yazuvannya zadach minimizaciyi sho ne ye opuklimi vidbuvayutsya v oblasti uzagalnenoyi opuklosti takozh vidomoyi yak abstraktnij opuklij analiz Div takozh RedaguvatiDvoyistist optimizaciya Umovi Karusha Kuna Takera Zadacha optimizaciyi Metod proksimalnogo gradiyenta Algoritm Frank VulfaPrimitki Redaguvati a b Nesterov ta Nemirovskii 1994 Murty Katta Kabadi Santosh 1987 Some NP complete problems in quadratic and nonlinear programming Mathematical Programming 39 2 117 129 doi 10 1007 BF02592948 Sahni S Computationally related problems in SIAM Journal on Computing 3 262 279 1974 Quadratic programming with one negative eigenvalue is NP hard Panos M Pardalos and Stephen A Vavasis in Journal of Global Optimization Volume 1 Number 1 1991 pg 15 22 a b v g d e zh i Boyd ta Vandenberghe 2004 Chritensen Klarbring chpt 4 Schmit L A Fleury C 1980 Structural synthesis by combining approximation concepts and dual methods Hiriart Urruty Jean Baptiste Lemarechal Claude 1996 Convex analysis and minimization algorithms Fundamentals s 291 ISBN 9783540568506 Ben Tal Aharon Nemirovskiĭ Arkadiĭ Semenovich 2001 Lectures on modern convex optimization analysis algorithms and engineering applications s 335 336 ISBN 9780898714913 Rockafellar R Tyrrell 1993 Lagrange multipliers and optimality SIAM Review 35 2 183 238 doi 10 1137 1035044 Proignorovano nevidomij parametr citeseerx dovidka Agrawal Akshay Verschueren Robin Diamond Steven Boyd Stephen 2018 A rewriting system for convex optimization problems Control and Decision 5 1 42 60 arXiv 1709 04494 doi 10 1080 23307706 2017 1397554 Dlya metodiv dlya opukloyi minimizaciyi div knigi vid Hiriart Urruty i Lemarechal a takozh pidruchniki vid Ruszczynski i Bertsekas i vid Boyd i Vandenberghe vnutrishnya tochka Nesterov Yurii Arkadii Nemirovskii 1995 Interior Point Polynomial Algorithms in Convex Programming Society for Industrial and Applied Mathematics ISBN 978 0898715156 Peng Jiming Roos Cornelis Terlaky Tamas 2002 Self regular functions and new search directions for linear and semidefinite optimization Mathematical Programming 93 1 129 171 ISSN 0025 5610 doi 10 1007 s101070200296 BertsekasSpisok literaturi RedaguvatiBertsekas Dimitri P Nedic Angelia Ozdaglar Asuman 2003 Convex Analysis and Optimization Belmont MA Athena Scientific ISBN 978 1 886529 45 8 Bertsekas Dimitri P 2009 Convex Optimization Theory Belmont MA Athena Scientific ISBN 978 1 886529 31 1 Bertsekas Dimitri P 2015 Convex Optimization Algorithms Belmont MA Athena Scientific ISBN 978 1 886529 28 1 Boyd Stephen P Vandenberghe Lieven 2004 Convex Optimization Cambridge University Press ISBN 978 0 521 83378 3 Procitovano 15 zhovtnya 2011 Borvejn Dzhonatan ta Lyuyis Adrian 2000 Analiz opuklosti ta nelinijna optimizaciya Springer Christensen Peter W Anders Klarbring 2008 An introduction to structural optimization 153 Springer Science amp Businees Media ISBN 9781402086663 Christensen Peter W Anders Klarbring 2008 An introduction to structural optimization 153 Springer Science amp Businees Media ISBN 9781402086663 Christensen Peter W Anders Klarbring 2008 An introduction to structural optimization 153 Springer Science amp Businees Media ISBN 9781402086663 Hiriart Urruti Zhan Batist i Lemareshal Klod 2004 Osnovi opuklogo analizu Berlin Springer Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume I Fundamentals Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 305 Berlin Springer Verlag s xviii 417 ISBN 978 3 540 56850 6 MR 1261420 Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume I Fundamentals Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 305 Berlin Springer Verlag s xviii 417 ISBN 978 3 540 56850 6 MR 1261420 Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume I Fundamentals Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 305 Berlin Springer Verlag s xviii 417 ISBN 978 3 540 56850 6 MR 1261420 Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume II Advanced theory and bundle methods Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 306 Berlin Springer Verlag s xviii 346 ISBN 978 3 540 56852 0 MR 1295240 Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume II Advanced theory and bundle methods Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 306 Berlin Springer Verlag s xviii 346 ISBN 978 3 540 56852 0 MR 1295240 Hiriart Urruty Jean Baptiste Lemarechal Claude 1993 Convex analysis and minimization algorithms Volume II Advanced theory and bundle methods Grundlehren der Mathematischen Wissenschaften Fundamental Principles of Mathematical Sciences 306 Berlin Springer Verlag s xviii 346 ISBN 978 3 540 56852 0 MR 1295240 Kiwiel Krzysztof C 1985 Methods of Descent for Nondifferentiable Optimization Lecture Notes in Mathematics New York Springer Verlag ISBN 978 3 540 15642 0 Kiwiel Krzysztof C 1985 Methods of Descent for Nondifferentiable Optimization Lecture Notes in Mathematics New York Springer Verlag ISBN 978 3 540 15642 0 Kiwiel Krzysztof C 1985 Methods of Descent for Nondifferentiable Optimization Lecture Notes in Mathematics New York Springer Verlag ISBN 978 3 540 15642 0 Lemarechal Claude 2001 Lagrangian relaxation U Michael Junger and Denis Naddef Computational combinatorial optimization Papers from the Spring School held in Schloss Dagstuhl May 15 19 2000 Lecture Notes in Computer Science 2241 Berlin Springer Verlag s 112 156 ISBN 978 3 540 42877 0 MR 1900016 doi 10 1007 3 540 45586 8 4 Lemarechal Claude 2001 Lagrangian relaxation U Michael Junger and Denis Naddef Computational combinatorial optimization Papers from the Spring School held in Schloss Dagstuhl May 15 19 2000 Lecture Notes in Computer Science 2241 Berlin Springer Verlag s 112 156 ISBN 978 3 540 42877 0 MR 1900016 doi 10 1007 3 540 45586 8 4 Lemarechal Claude 2001 Lagrangian relaxation U Michael Junger and Denis Naddef Computational combinatorial optimization Papers from the Spring School held in Schloss Dagstuhl May 15 19 2000 Lecture Notes in Computer Science 2241 Berlin Springer Verlag s 112 156 ISBN 978 3 540 42877 0 MR 1900016 doi 10 1007 3 540 45586 8 4 Nesterov Yurii Nemirovskii Arkadii 1994 Interior Point Polynomial Methods in Convex Programming SIAM Nesterov Yurij 2004 Vstupni lekciyi z opukloyi optimizaciyi naukovi vidavci Kluwer Rockafellar R T 1970 Convex analysis Princeton Princeton University Press Ruszczynski Andrzej 2006 Nonlinear Optimization Princeton University Press Shmit L A Fleri C 1980 Strukturnij sintez shlyahom poyednannya koncepcij nablizhennya ta podvijnih metodiv Dzh Amer Inst Aeronavt Astronavt 18 1252 1260Posilannya RedaguvatiStiven Bojd ta Liven Vandenberge opukla optimizaciya kniga v pdf EE364a Opukla optimizaciya I ta EE364b Opukla optimizaciya II domashni storinki kursu Stenford 6 253 Opuklij analiz ta optimizaciya domashnya storinka kursu MIT OCW Brajan Borchers Oglyad programnogo zabezpechennya dlya opukloyi optimizaciyi Otrimano z https uk wikipedia org w index php title Opukla optimizaciya amp oldid 39706965