{"id":{"repo_id":"ets-quebec","oai_identifier":"oai:espace.etsmtl.ca:2992"},"canonical_url":"https://search.ndltd.org/etd/ets-quebec/oai:espace.etsmtl.ca:2992","repository":{"repo_id":"ets-quebec","name":"ETS (Quebec)","base_url":"https://espace.etsmtl.ca/cgi/oai2"},"display":{"title":"Adaptive client selection and upgrade of resources for robust federated learning","abstract":"Driven by privacy concerns and the visions of Deep Learning, the last four years have witnessed a paradigm shift in the applicability mechanism of Machine Learning (ML). An emerging model, called Federated Learning (FL), is rising above both centralized systems and on-site analysis, to be a new fashioned design for ML implementation. It is a privacy preserving decentralized approach, which keeps raw data on devices and involves local ML training while eliminating data communication overhead. A federation of the learned and shared models is then performed on a central server to aggregate and share the built knowledge among participants. Recently, many research interests have been drawn targeting different aspects in FL. However, the limitations while selecting the clients and reducing the communication costs still threaten FL performance and its applicability in real scenarios. From these premises, this dissertation fills the lacking of the existing research outcomes. Particularly, the contributions of this thesis are threefold: (1) a proof-of-concept to leverage FL for IoT security, where FL based scheme is proposed for IoT intrusion detection showing its outperformance across the centralized and on-device learning (2) multi-criteria client selection for optimal IoT FL, where the number of clients is maximized in each round and the clients resources are studied to predict those able to successfully complete the training task and (3) adaptive upgrade of client resources for improving the quality of FL model, where the contributions of the clients updates are measured and accordingly available resources are allocated among the clients to benefit from their high-quality data and to reach the target global model performance. The efficiency and robustness of the proposed approaches have been proved through various prototypes and thorough experiments.","abstract_html":"Driven by privacy concerns and the visions of Deep Learning, the last four years have witnessed a paradigm shift in the applicability mechanism of Machine Learning (ML). An emerging model, called Federated Learning (FL), is rising above both centralized systems and on-site analysis, to be a new fashioned design for ML implementation. It is a privacy preserving decentralized approach, which keeps raw data on devices and involves local ML training while eliminating data communication overhead. A federation of the learned and shared models is then performed on a central server to aggregate and share the built knowledge among participants. Recently, many research interests have been drawn targeting different aspects in FL. However, the limitations while selecting the clients and reducing the communication costs still threaten FL performance and its applicability in real scenarios. From these premises, this dissertation fills the lacking of the existing research outcomes. Particularly, the contributions of this thesis are threefold: (1) a proof-of-concept to leverage FL for IoT security, where FL based scheme is proposed for IoT intrusion detection showing its outperformance across the centralized and on-device learning (2) multi-criteria client selection for optimal IoT FL, where the number of clients is maximized in each round and the clients resources are studied to predict those able to successfully complete the training task and (3) adaptive upgrade of client resources for improving the quality of FL model, where the contributions of the clients updates are measured and accordingly available resources are allocated among the clients to benefit from their high-quality data and to reach the target global model performance. The efficiency and robustness of the proposed approaches have been proved through various prototypes and thorough experiments.","abstract_has_math":false,"creators":["Abdul Rahman, Sawsan"],"institution":"École de technologie supérieure","degree_name":"Thèse de doctorat électronique","degree_level":"phd","degree_discipline":null,"degree_department":null,"school":null,"contributors":[],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-03-11","date_published":"2022-03-11","updated_at":"2026-11-18T22:09:26Z","subjects":["apprentissage fédéré, internet des objets, selection multicritère, optimisation à deux niveaux, allocation des ressources, kubernetes"],"languages":["en"],"rights":[],"rights_urls":[],"identifier_entries":[{"key":"dc:identifier","label":"Identifier","values":["INSERT-YOUR-LIBRARY-ID-HERE2992"],"render_values":[{"text":"INSERT-YOUR-LIBRARY-ID-HERE2992","href":null,"code":true}]}],"person_identifier_entries":[]},"links":{"outbound_url":null,"outbound_label":null,"outbound_source":null,"record_url":null,"record_label":null,"record_kind":null,"record_source":null,"download_url":"https://espace.etsmtl.ca/id/eprint/2992/1/ABDUL_RAHMAN_Sawsan.pdf","download_label":"Download PDF","download_source":"dc:identifier","access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://espace.etsmtl.ca/cgi/oai2?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Aespace.etsmtl.ca%3A2992","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:creator","label":"Author","values":["Abdul Rahman, Sawsan"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2022-03-11"]},{"key":"dc:publisher","label":"Institution","values":["École de technologie supérieure"]},{"key":"dc:type","label":"Dc Type","values":["Electronic Thesis or Dissertation"]},{"key":"thesis:degree_level","label":"Degree Level","values":["phd"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Thèse de doctorat électronique"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["École de technologie supérieure"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["apprentissage fédéré, internet des objets, selection multicritère, optimisation à deux niveaux, allocation des ressources, kubernetes"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["INSERT-YOUR-LIBRARY-ID-HERE2992","https://espace.etsmtl.ca/id/eprint/2992/1/ABDUL_RAHMAN_Sawsan.pdf"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Driven by privacy concerns and the visions of Deep Learning, the last four years have witnessed a paradigm shift in the applicability mechanism of Machine Learning (ML). An emerging model, called Federated Learning (FL), is rising above both centralized systems and on-site analysis, to be a new fashioned design for ML implementation. It is a privacy preserving decentralized approach, which keeps raw data on devices and involves local ML training while eliminating data communication overhead. A federation of the learned and shared models is then performed on a central server to aggregate and share the built knowledge among participants. Recently, many research interests have been drawn targeting different aspects in FL. However, the limitations while selecting the clients and reducing the communication costs still threaten FL performance and its applicability in real scenarios. From these premises, this dissertation fills the lacking of the existing research outcomes. Particularly, the contributions of this thesis are threefold: (1) a proof-of-concept to leverage FL for IoT security, where FL based scheme is proposed for IoT intrusion detection showing its outperformance across the centralized and on-device learning (2) multi-criteria client selection for optimal IoT FL, where the number of clients is maximized in each round and the clients resources are studied to predict those able to successfully complete the training task and (3) adaptive upgrade of client resources for improving the quality of FL model, where the contributions of the clients updates are measured and accordingly available resources are allocated among the clients to benefit from their high-quality data and to reach the target global model performance. The efficiency and robustness of the proposed approaches have been proved through various prototypes and thorough experiments."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Adaptive client selection and upgrade of resources for robust federated learning"]}]}],"canonical_facts":{"dc:creator":["Abdul Rahman, Sawsan"],"dc:date":["2022-03-11"],"dc:description":["Driven by privacy concerns and the visions of Deep Learning, the last four years have witnessed a paradigm shift in the applicability mechanism of Machine Learning (ML). An emerging model, called Federated Learning (FL), is rising above both centralized systems and on-site analysis, to be a new fashioned design for ML implementation. It is a privacy preserving decentralized approach, which keeps raw data on devices and involves local ML training while eliminating data communication overhead. A federation of the learned and shared models is then performed on a central server to aggregate and share the built knowledge among participants. Recently, many research interests have been drawn targeting different aspects in FL. However, the limitations while selecting the clients and reducing the communication costs still threaten FL performance and its applicability in real scenarios. From these premises, this dissertation fills the lacking of the existing research outcomes. Particularly, the contributions of this thesis are threefold: (1) a proof-of-concept to leverage FL for IoT security, where FL based scheme is proposed for IoT intrusion detection showing its outperformance across the centralized and on-device learning (2) multi-criteria client selection for optimal IoT FL, where the number of clients is maximized in each round and the clients resources are studied to predict those able to successfully complete the training task and (3) adaptive upgrade of client resources for improving the quality of FL model, where the contributions of the clients updates are measured and accordingly available resources are allocated among the clients to benefit from their high-quality data and to reach the target global model performance. The efficiency and robustness of the proposed approaches have been proved through various prototypes and thorough experiments."],"dc:format":["application/pdf"],"dc:identifier":["INSERT-YOUR-LIBRARY-ID-HERE2992","https://espace.etsmtl.ca/id/eprint/2992/1/ABDUL_RAHMAN_Sawsan.pdf"],"dc:language":["en"],"dc:publisher":["École de technologie supérieure"],"dc:subject":["apprentissage fédéré, internet des objets, selection multicritère, optimisation à deux niveaux, allocation des ressources, kubernetes"],"dc:title":["Adaptive client selection and upgrade of resources for robust federated learning"],"dc:type":["Electronic Thesis or Dissertation"],"thesis:degree_level":["phd"],"thesis:degree_name":["Thèse de doctorat électronique"],"thesis:institution_name":["École de technologie supérieure"]},"updated_at":"2026-11-18T22:09:26Z"}