Vis enkel innførsel

dc.contributor.authorSamer, Phillippe
dc.date.accessioned2023-12-19T08:32:13Z
dc.date.available2023-12-19T08:32:13Z
dc.date.issued2023-12-21
dc.date.submitted2023-11-20T15:25:00.222Z
dc.identifiercontainer/87/e3/f8/33/87e3f833-4407-4930-ab2b-5401bf99014b
dc.identifier.isbn9788230863671
dc.identifier.isbn9788230841396
dc.identifier.urihttps://hdl.handle.net/11250/3108108
dc.description.abstractI denne avhandlinga interesserer vi oss for å finne delgrafer som svarer til utvalgte modeller for begrepene sammenheng og uavhengighet. I korthet betyr dette stabile (også kalt uavhengige) mengder med gitt kardinalitet, stabile (også kalt konfliktfrie) spenntrær og pardannelser (eller uavhengige kantmengder) som induserer en sammenhengende delgraf. Dette er kombinatoriske strukturer som kan generaliseres til ulike modeller for nettverksdesign innen telekommunikasjon og forsyningsvirksomhet, plassering av anlegg, fylogenetikk, og mange andre applikasjoner innen operasjonsanalyse og optimering. Vi argumenterer for at de valgte strukturene reiser interessante forskningsspørsmål, og vi bidrar med forbedret matematisk forståelse av selve strukturene, samt forbedrede algoritmer for å takle de tilhørende kombinatoriske optimeringsproblemene. Med det mener vi metoder for å identifisere en optimal struktur, forutsatt at elementene som danner dem (hjørner eller kanter i en gitt graf) er tildelt verdier. Forskninga vår omfatter ulike områder innenfor algoritmer, kombinatorikk og optimering. De fleste resultatene omhandler det å finne bedre beskrivelser av de geometriske strukturene (nemlig 0/1-polytoper) som representerer alle mulige løsninger for hvert av problemene. Slike forbedrede beskrivelser oversettes til lineære ulikheter i heltallsprogrammeringsmodeller, noe som igjen gir mer effektive beregningsresultater når man løser referanseinstanser av hvert problem. Vi påpeker gjentatte ganger betydninga av å dele kildekoden til implementasjonen av alle utviklede algoritmer og verktøy når det foreslås nye modeller og løsningsmetoder for heltallsprogrammering og kombinatorisk optimering. Kodearkivene våre inkluderer fullstendige implementasjoner, utformet med effektivitet og modulær design i tankene, og fremmer dermed gjenbruk, videre forskning og nye anvendelser innen forskning og utvikling.en_US
dc.description.abstractWe are interested in finding subgraphs that capture selected models of connectivity and independence. In short: fixed cardinality stable (or independent) sets, stable (or conflict-free) spanning trees, and matchings (or independent edge sets) inducing a connected subgraph. These are combinatorial structures that can be generalized to a number of models across network design in telecommunication and utilities, facility location, phylogenetics, among many other application domains of operations research and optimization. We argue that the selected structures raise appealing research questions, and seek to contribute with improved mathematical understanding of the structures themselves, as well as improved algorithms to face the corresponding combinatorial optimization problems. That is, methods to identify an optimal structure, assuming the elements that form them (vertices or edges in a given graph) have a weight. Our research spans different lines within algorithmics, combinatorics and optimization. Most of the results concern finding better descriptions of the geometric structures (namely, 0/1-polytopes) that represent all feasible solutions to each of the problems. Such improved descriptions translate to linear inequalities in integer programming formulations which, in turn, provide stronger computational results when solving benchmark instances of each problem. We repeatedly remark the importance of sharing an open-source implementation of all algorithms and tools developed when proposing new models and solution methods in integer programming and combinatorial optimization. Our code repositories include full implementations, crafted with efficiency and modular design in mind, thus fostering reuse, further research and new applications in research and development.en_US
dc.language.isoengen_US
dc.publisherThe University of Bergenen_US
dc.relation.haspartPaper 1: Phillippe Samer, Dag Haugland, The unsuitable neighbourhood inequalities for the fixed cardinality stable set polytope, chapter in Graphs and Combinatorial Optimization: from theory to applications, AIRO Springer Series, Volume 5, 2021. The accepted version is available at: <a href="https://hdl.handle.net/11250/2761120" target="blank"> https://hdl.handle.net/11250/2761120</a>en_US
dc.relation.haspartPaper 2: Phillippe Samer, Dag Haugland, Fixed cardinality stable sets, Discrete Applied Mathematics, Volume 303, 2021. The article is available at: <a href="https://hdl.handle.net/11250/2935773" target="blank">https://hdl.handle.net/11250/2935773</a>en_US
dc.relation.haspartPaper 3: Phillippe Samer, Dag Haugland, Towards stronger Lagrangean bounds for stable spanning trees, peer-reviewed proceedings of The 10th International Network Optimization Conference, Aachen, Germany, 2022. The chapter is available at: <a href="https://hdl.handle.net/11250/3021704" target="blank">https://hdl.handle.net/11250/3021704</a>en_US
dc.relation.haspartPaper 4: Phillippe Samer, Dag Haugland, Polyhedral results and stronger Lagrangean bounds for stable spanning trees, Optimization Letters, Volume 17, Issue 6, 2023. The article is available at: <a href="https://hdl.handle.net/11250/3038498" target="blank">https://hdl.handle.net/11250/3038498</a>en_US
dc.relation.haspartPaper 5: Phillippe Samer, Phablo F.S. Moura, Polyhedral approach to weighted connected matchings in general graphs. The article is not available in BORA. The submitted version is available at: <a href="https://doi.org/10.48550/arXiv.2310.05733" target="blank">https://doi.org/10.48550/arXiv.2310.05733</a>en_US
dc.relation.haspartPaper 6: Phillippe Samer, On a class of strong valid inequalities for the connected matching Polytope. The article is not available in BORA. The submitted version is available at: <a href=" https://doi.org/10.48550/arXiv.2309.14019" target="blank">https://doi.org/10.48550/arXiv.2309.14019</a>en_US
dc.rightsAttribution-NonCommercial (CC BY-NC). This item's rights statement or license does not apply to the included articles in the thesis.
dc.rights.urihttps://creativecommons.org/licenses/by-nc/4.0/
dc.titlePolyhedra and algorithms for problems bridging notions of connectivity and independenceen_US
dc.typeDoctoral thesisen_US
dc.date.updated2023-11-20T15:25:00.222Z
dc.rights.holderCopyright the Author.en_US
dc.contributor.orcid0000-0001-9007-0237
dc.description.degreeDoktorgradsavhandling
fs.unitcode12-12-0


Tilhørende fil(er)

Thumbnail

Denne innførselen finnes i følgende samling(er)

Vis enkel innførsel

Attribution-NonCommercial (CC BY-NC). This item's rights statement or license does not apply to the included articles in the thesis.
Med mindre annet er angitt, så er denne innførselen lisensiert som Attribution-NonCommercial (CC BY-NC). This item's rights statement or license does not apply to the included articles in the thesis.