[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-126048-en":3,"doc-seo-126048-105":31,"detail-sidebar-cat-0-en-105":93},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":20,"is_deleted":4,"is_public":21,"is_downloadable":21,"audit_status":21,"page_count":22,"language":23,"language_code":24,"site_id":25,"html_lang":24,"table_of_contents":26,"faqs":27,"seo_title":28,"seo_description":14,"update_tm":29,"read_time":30},126048,962084928904,"Asher","https://ap-avatar.wpscdn.com/davatar_155a257f0dc6eb9ab79c44ca47cae57d",8,"Research & Report","Graph Signal Processing techniques for Machine Learning optimization - Master’s Thesis","Graph Signal Processing (GSP) provides a flexible way to extend classical signal processing tools such as filtering and sampling from regular domains to irregular structures including social networks. Within this framework, the Q-GFT generalizes the Graph Fourier Transform by introducing a non-trivial inner product and enabling greater flexibility via its variation operator, graph partitioning strategies, and changes to spectral properties. The thesis integrates Q-GFT with graph neural networks for node classification, analyzing effects on GCN and GraphSAGE and linking energy concentration patterns to model performance.","Graph Signal Processing techniques for Machine Learning optimization  \nMaster’s Thesis submitted to the Faculty of the Escola T`ecnica d’Enginyeria de Telecomunicaci´o de Barcelona  \nUniversitat Polit`ecnica de Catalunya by  \nPablo Beaus Iranzo  \nIn partial fulfillment  \nof the requirements for the Master’s degree in Advanced Telecommunication Technologies  \nAdvisors:  \nFerran Marqu´es  \nAntonio Ortega  \nAjitesh Srivastava  \nBarcelona, October 2024  \nAbstract  \nGraph Signal Processing (GSP) offers a flexible framework for extending classical signal processing techniques, such as filtering and sampling, to irregular domains like social networks. In this context, the Q-GFT, a generalization of the Graph Fourier Transform (GFT) that incorporates a non-trivial inner product, was introduced to provide a more flexible analysis of graph signals. The Q-GFT brings additional flexibility through its variation operator, graph partitioning strategies, and modification of spectral properties.  \nIn this thesis, we explore the integration of the Q-GFT with graph neural networks (GNNs) for node classification tasks. We investigate the impact of Q-GFT and evaluate their effect on well-known GNN architectures such as Graph Convolutional Networks (GCNs) and GraphSAGE. Results provide valuable insights into the interaction between graph partitioning strategies and neural networks, showing how energy concentration patterns from different Q-based Graph Fourier Transforms (Q-GFT) correlate with model performance. While most models did not outperform the baseline, using ground-truth labels for the Fully Balanced max-cut did, suggesting potential for new partitioning methods to advance graph machine learning.  \nResum  \nEl processament de senyals en grafs (GSP) ofereix un marc flexible per a estendre lest`ecniques cl`assiques de processament de senyals, com el filtratge i el mostreig, a dominisirregulars com les xarxes socials. En aquest context, es va introduir la Q-GFT, una generalitzaci´o de la Transformada de Fourier de Grafs (GFT) que incorpora un producte intern no trivial, per a proporcionar una an`alisi m´es flexible dels senyals de grafs. La Q-GFTaporta flexibilitat addicional a trav´es del seu operador de variaci´o, estrat`egies de partici´ode grafs i modificaci´o de les propietats espectrals.  \nEn aquesta tesi, explorem la integraci´o de la Q-GFT amb xarxes neuronals de grafs (GNNs) per a tasques de classificaci´o de nodes. Investiguem l’impacte de la Q-GFT iavaluem el seu efecte en arquitectures de GNN ben conegudes com les xarxes convolucionals de grafs (GCNs) i GraphSAGE. Els resultats proporcionen informaci´o valuosa sobre la interacci´o entre les estrat`egies de partici´o de grafs i les xarxes neuronals, mostrant com els patrons de concentraci´o d’energia de diferents Transformades de Fourier de Grafsbasades en Q-GFT es correlacionen amb el rendiment del model. Encara que la majoria dels models no superin la refer`encia de base, l’´us d’etiquetes reals per a un tall m`axim totalment equilibrat s´ı que l’aconsegueix, suggerint potencial de nous m`etodes de partici´oper a avan¸car en l’aprenentatge autom`atic de grafs.  \nResumen  \nEl procesado de se˜nal en grafos (GSP) ofrece un marco flexible para extender las t´ecnicascl´asicas de procesado de se˜nal, como el filtrado y el muestreo, a dominios irregulares como las redes sociales. En este contexto, se introdujo la Q-GFT, una generalizaci´on de la Transformada de Fourier de Grafos (GFT) que incorpora un producto interno no trivial, para proporcionar un an´alisis m´as flexible de las se˜nales de grafos. La Q-GFT aporta flexibilidad adicional a trav´es de su operador de variaci´on, estrategias de partici´on de grafos y modificaci´on de las propiedades espectrales.  \nEn esta tesis, exploramos la integraci´on de la Q-GFT con redes neuronales de grafos (GNNs) para tareas de clasificaci´on de nodos. Investigamos el impacto de la Q-GFT y evaluamos su efecto en arquitecturas de GNN bien conocidas","cbCaip1iY1JYOZlE","https://ap.wps.com/l/cbCaip1iY1JYOZlE","pdf",3381872,11,1,44,"English","en",105,"# Introduction\n## Project overview, motivation and goals\n## Structure\n# Background and State of the Art Technology\n## Graphs and Graph Signal Processing\n## I","[{\"question\":\"What is the role of Q-GFT in graph signal processing?\",\"answer\":\"Q-GFT generalizes the Graph Fourier Transform by using a non-trivial inner product, offering more flexible analysis through a variation operator, partitioning strategies, and altered spectral properties.\"},{\"question\":\"How does the thesis apply Q-GFT to machine learning models?\",\"answer\":\"It integrates Q-GFT with graph neural networks to perform node classification, evaluating impacts on architectures such as GCNs and GraphSAGE.\"},{\"question\":\"What do the results reveal about graph partitioning and performance?\",\"answer\":\"Energy concentration patterns produced by different Q-based Graph Fourier Transforms correlate with model performance, and using ground-truth labels for a fully balanced max-cut shows improvement over the baseline.\"}]","Graph Signal Processing techniques for Machine Learning optimization - Master’s Thesis | PDF",1785902767,111,{"code":4,"msg":32,"data":33},"ok",{"site_id":25,"language":24,"slug":34,"title":13,"keywords":35,"description":14,"schema_data":36,"social_meta":88,"head_meta":90,"extra_data":92,"updated_unix":29},"graph-signal-processing-techniques-for-machine-learning-optimization-masters-thesis","",{"@graph":37,"@context":87},[38,55,70],{"@type":39,"itemListElement":40},"BreadcrumbList",[41,45,49,52],{"item":42,"name":43,"@type":44,"position":21},"https://docshare.wps.com","Home","ListItem",{"item":46,"name":47,"@type":44,"position":48},"https://docshare.wps.com/document/","Document",2,{"item":50,"name":12,"@type":44,"position":51},"https://docshare.wps.com/document/research-report/",3,{"item":53,"name":13,"@type":44,"position":54},"https://docshare.wps.com/document/graph-signal-processing-techniques-for-machine-learning-optimization-masters-thesis/126048/",4,{"url":53,"name":13,"@type":56,"author":57,"headline":13,"publisher":59,"fileFormat":62,"inLanguage":24,"description":14,"dateModified":63,"datePublished":64,"encodingFormat":62,"isAccessibleForFree":65,"interactionStatistic":66},"DigitalDocument",{"name":9,"@type":58},"Person",{"url":42,"name":60,"@type":61},"DocShare","Organization","application/pdf","2026-08-23","2026-08-05",true,{"@type":67,"interactionType":68,"userInteractionCount":20},"InteractionCounter",{"@type":69},"ViewAction",{"@type":71,"mainEntity":72},"FAQPage",[73,79,83],{"name":74,"@type":75,"acceptedAnswer":76},"What is the role of Q-GFT in graph signal processing?","Question",{"text":77,"@type":78},"Q-GFT generalizes the Graph Fourier Transform by using a non-trivial inner product, offering more flexible analysis through a variation operator, partitioning strategies, and altered spectral properties.","Answer",{"name":80,"@type":75,"acceptedAnswer":81},"How does the thesis apply Q-GFT to machine learning models?",{"text":82,"@type":78},"It integrates Q-GFT with graph neural networks to perform node classification, evaluating impacts on architectures such as GCNs and GraphSAGE.",{"name":84,"@type":75,"acceptedAnswer":85},"What do the results reveal about graph partitioning and performance?",{"text":86,"@type":78},"Energy concentration patterns produced by different Q-based Graph Fourier Transforms correlate with model performance, and using ground-truth labels for a fully balanced max-cut shows improvement over the baseline.","https://schema.org",{"og:url":53,"og:type":89,"og:title":13,"og:site_name":60,"og:description":14},"article",{"robots":91,"canonical":53},"index,follow",{"doc_id":7,"site_id":25},{"code":4,"msg":5,"data":94},[95,99,103,107,112,117,122,125,130,133,137],{"id":21,"doc_module":4,"doc_module_name":47,"category_name":96,"show_sort_weight":97,"slug":98},"Story & Novel",90,"story-novel",{"id":48,"doc_module":4,"doc_module_name":47,"category_name":100,"show_sort_weight":101,"slug":102},"Literature",80,"literature",{"id":54,"doc_module":4,"doc_module_name":47,"category_name":104,"show_sort_weight":105,"slug":106},"Exam",70,"exam",{"id":108,"doc_module":4,"doc_module_name":47,"category_name":109,"show_sort_weight":110,"slug":111},5,"Comic",60,"comic",{"id":113,"doc_module":4,"doc_module_name":47,"category_name":114,"show_sort_weight":115,"slug":116},6,"Technology",50,"technology",{"id":118,"doc_module":4,"doc_module_name":47,"category_name":119,"show_sort_weight":120,"slug":121},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":47,"category_name":12,"show_sort_weight":123,"slug":124},30,"research-report",{"id":126,"doc_module":4,"doc_module_name":47,"category_name":127,"show_sort_weight":128,"slug":129},9,"Religion & Spirituality",20,"religion-spirituality",{"id":128,"doc_module":4,"doc_module_name":47,"category_name":131,"show_sort_weight":128,"slug":132},"World Cup","world-cup",{"id":134,"doc_module":4,"doc_module_name":47,"category_name":135,"show_sort_weight":134,"slug":136},10,"Lifestyle","lifestyle",{"id":138,"doc_module":4,"doc_module_name":47,"category_name":139,"show_sort_weight":108,"slug":140},19,"General","general"]