Phd Position F/m Topology Design for Decentralized
Il y a 5 mois
Le descriptif de l’offre ci-dessous est en Anglais_
**Type de contrat **:CDD
**Niveau de diplôme exigé **:Bac + 5 ou équivalent
**Fonction **:Doctorant
**A propos du centre ou de la direction fonctionnelle**:
The Inria centre at Université Côte d'Azur includes 37 research teams and 8 support services. The centre's staff (about 500 people) is made up of scientists of different nationalities, engineers, technicians and administrative staff. The teams are mainly located on the university campuses of Sophia Antipolis and Nice as well as Montpellier, in close collaboration with research and higher education laboratories and establishments (Université Côte d'Azur, CNRS, INRAE, INSERM...), but also with the regiona economic players.
With a presence in the fields of computational neuroscience and biology, data science and modeling, software engineering and certification, as well as collaborative robotics, the Inria Centre at Université Côte d'Azur is a major player in terms of scientific excellence through its results and collaborations at both European and international levels.
**Contexte et atouts du poste**:
The research activity will be supervised by
**Mission confiée**:
Topology Design for Decentralized Federated Learning
# Context
In the classic FL setting, a server coordinates the training phase. At each training round, the server sends the current model to the clients, which individually train on their local datasets and send model updates to the server, which in turn aggregates them (often through a simple averaging operation). In contrast to this client-server approach, decentralized FL algorithms (also called P2P FL algorithms) work by having each client communicate directly with a subset of the clients (its neighbours): this process alternates between model updates and weighted averaging of the neighbours' models (consensus-based optimization). Decentralized algorithms can take advantage of good pairwise connectivity, avoid the potential communication bottleneck at the server [marfoq20] as well as provide better privacy guarantees [cyffers22].
The communication graph (i.e., the graph induced by clients' pairwise communications) and the local clients' aggregation strategies play a fundamental role in determining FL algorithms' convergence speed. In particular, the communication topology has two contrasting effects on training time. First, a more connected topology leads to faster convergence in terms of number of communication rounds [nedic18]. Second, a more connected topology increases the duration of a communication round (e.g., because it may cause network congestion), motivating the use of degree-bounded topologies where every client sends and receives a small number of messages at each round [lian17]. Most of the existing literature has focused on one aspect or the other.
The classic literature on consensus-based optimization has quantified the effect of the communication topology on the number of rounds through worst-case convergence bounds in terms of the spectral gap of the consensus matrix (i.e., the matrix with the averaging weight), see [nedic18] and references there. Later papers have highlighted the convergence rate' insensitivity to the spectral gap for a large number of communication rounds and small learning rates [lian17,koloskova21,pu20].
Another line of work has shown that the effect of the topology is less important if local data distributions [neglia20] or average data distributions in each neighborhood [lebars23,dandi22] are close to the average data distribution over the whole population. In the extreme case of homogeneous local distributions, one may even prefer consensus matrices with poor spectral properties because they enable the use of larger learning rates [vogel22].
A separate line of works has studied how to design the communication topology in order to minimize the duration of one round, taking into account the variability of the computation times [neglia19] or the characteristics of Internet connections [marfoq20].
# Research objectives
The goal of this PhD is to propose algorithms to design the communication topology for decentralized federated learning with the goal of minimizing the total training duration, taking into account how connectivity affect both the number of rounds required and the duration of a single round.
Several settings will be considered: in particular, one may construct the topology in a pre-processing step (prior to learning), or dynamically while learning. Dynamic topology design can be a way to tackle online decentralized learning [asadi22,marfoq23], where the topology is adjusted and refined as clients collect more data.
Finally, he/she will also study to what extent the existing results can be extended to asymmetric communication links and other distributed optimization algorithms like push-sum ones [kempe03,benezit10].
# References
[asadi22] M. Asadi, A. Bellet, O.A. Maillard and M. Tommasi. Collaborative Algorith
-
PhD Position in Topology Design for Decentralized Federated Learning
il y a 3 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinJob DetailsThis PhD research position is part of the INRIA research initiative on Federated Learning, FedMalin. The successful candidate will join the NEO project-team, which is positioned at the intersection of Operations Research and Network Science. The research activity will be supervised by Giovanni Neglia, Aurélien Bellet.Research ObjectivesThe goal...
-
PhD Position in Topology Design for Decentralized Federated Learning
il y a 4 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinResearch OpportunityThis PhD thesis is part of Inria's research initiative on Federated Learning, FedMalin. The PhD candidate will join the NEO project-team, positioned at the intersection of Operations Research and Network Science.Research ObjectivesThe goal of this PhD is to propose algorithms to design the communication topology for decentralized...
-
PhD Position in Topology Design for Decentralized Federated Learning
il y a 4 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinResearch InitiativeThis PhD thesis is part of Inria's research initiative on Federated Learning, FedMalin. The PhD candidate will join the NEO project-team, positioned at the intersection of Operations Research and Network Science.Research ObjectivesThe goal of this PhD is to propose algorithms to design the communication topology for decentralized federated...
-
Reliability Expert for Decentralized Communication Systems
il y a 5 jours
Technopole de Sophia Antipolis, France INRIA Temps pleinDecentralized Communication Systems ResearcherThis is a postdoctoral research position at INRIA, focusing on the reliability and throughput of decentralized communication systems.
-
Technopole de Sophia Antipolis, France INRIA Temps pleinJob Title: Postdoctoral Researcher in Topology-Infused Generative Models for Galaxy Map ApproximationJob Description:The successful candidate will work on developing mathematical frameworks and practical setups for applying Topological Data Analysis (TDA) methods, descriptors, and features in the context of deep generative models. A particular focus will be...
-
Research Scientist
il y a 2 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinJob OfferINRIA is seeking a highly skilled Research Scientist to join our team in the development of novel topology-infused generative models for approximate galaxy map simulations. The successful candidate will work in the context of Topological Data Analysis (TDA) and Deep Generative Models (DGM), with a focus on multimodal variational autoencoders and...
-
Optimization Researcher PhD Position
il y a 2 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinProject OverviewThe position of PhD researcher in optimization research is part of a collaborative project between INRIA and the ANR SWAG-P project. The project focuses on the development of reduced-order modeling techniques for the robust design of gap plasmon resonators.The successful candidate will work on the development of novel optimization methods for...
-
Technopole de Sophia Antipolis, France INRIA Temps pleinProject OverviewThe proposed PhD project is part of a collaborative multidisciplinary project between the Atlantis project-team from INRIA Research Center and the Elena team at the Pascal Institute. The project aims to design efficient nanoscale biosensors using plasmonic nano-resonators, which are promising for their high sensitivity and miniaturization...
-
Postdoctoral Researcher
il y a 4 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinResearch Position OverviewThe INRIA research center is seeking a highly motivated postdoctoral researcher to join our team and contribute to the development of decentralized communications systems. The successful candidate will work in the broad intersection of information theory, communication theory, and game theory, with a focus on studying the...
-
Phd Thesis
Il y a 3 mois
Sophia Antipolis, France Laboratoire I3S - CNRS (UMR 7271) Temps plein**PhD Thesis - Error-Correction Codes for Reliable DNA Storage**: - Réf **ABG-125677** - Sujet de Thèse- 05/09/2024- Contrat doctoral- Laboratoire I3S - CNRS (UMR 7271)- Lieu de travail- Sophia Antipolis - Provence-Alpes-Côte d'Azur - France- Intitulé du sujet- PhD Thesis - Error-Correction Codes for Reliable DNA Storage- Champs scientifiques- Science...
-
Postdoctoral Researcher
il y a 4 semaines
Technopole de Sophia Antipolis, France INRIA Temps pleinResearch Position OverviewThe INRIA research center is seeking a highly skilled postdoctoral researcher to join our team and contribute to the development of decentralized communications systems. This position is part of a national collaboration network that includes INSA de Rennes and is supported by the French National Agency for Research (ANR) via the...
-
Phd Position F/m Privacy On-demand and Security
Il y a 5 mois
Sophia Antipolis, France Inria Temps pleinLe descriptif de l’offre ci-dessous est en Anglais_ **Type de contrat **:CDD**Niveau de diplôme exigé **:Bac + 5 ou équivalent**Fonction **:DoctorantA propos du centre ou de la direction fonctionnelle Le centre Inria d'Université Côte d'Azur regroupe 37 équipes de recherche et 8 services d’appui. Le personnel du centre (500 personnes environ) est...
-
Phd Position F/m Phd Position Computer Vision
Il y a 5 mois
Sophia Antipolis, France Inria Temps pleinLe descriptif de l’offre ci-dessous est en Anglais_ **Type de contrat **:CDD **Niveau de diplôme exigé **:Bac + 5 ou équivalent **Fonction **:Doctorant **A propos du centre ou de la direction fonctionnelle**: The Inria Université Côte d’Azur center counts 36 research teams as well as 7 support departments. The center's staff (about 500 people...
-
Technopole de Sophia Antipolis, France INRIA Temps pleinResearch Position at INRIAThe Centre Inria d'Université Côte d'Azur is seeking a highly skilled Postdoctoral Researcher to join our team and contribute to the development of topology-infused generative models for galaxy map approximation.Context and RequirementsThe successful candidate will work under the supervision of Mathieu Carrière and...
-
Phd Position F/m Generative Ai for The
Il y a 2 mois
Sophia Antipolis, France Inria Temps pleinLe descriptif de l’offre ci-dessous est en Anglais_ **Type de contrat **:CDD **Niveau de diplôme exigé **:Bac + 5 ou équivalent **Fonction **:Doctorant **A propos du centre ou de la direction fonctionnelle**: The Inria center at Université Côte d'Azur includes 42 research teams and 9 support services. The center’s staff (about 500 people) is...
-
Technopole de Sophia Antipolis, France INRIA Temps pleinProject OverviewThe INRIA Research Center at Université Côte d'Azur and Institut Pascal at Université Clermont Auvergne are seeking a PhD candidate to work on a collaborative multidisciplinary project focused on the robust design of gap plasmon resonators. This project is part of the ANR SWAG-P project, which aims to develop efficient numerical methods...
-
Phd Position F/m Monitoring Plane for Mobile
Il y a 5 mois
Sophia Antipolis, France Inria Temps pleinLe descriptif de l’offre ci-dessous est en Anglais_ **Type de contrat **:CDD**Niveau de diplôme exigé **:Bac + 5 ou équivalent**Fonction **:DoctorantA propos du centre ou de la direction fonctionnelle The Inria Université Côte d'Azur center counts 37 research teams as well as 8 support services. The center's staff (about 500 people) is made up of...
-
Technopole de Sophia Antipolis, France INRIA Temps pleinJob DescriptionThe INRIA is seeking a highly motivated postdoctoral researcher to work on the development of topology-infused generative models for the fast generation of approximate galaxy maps. The successful candidate will work under the supervision of Mathieu Carrière and will be located at the Centre Inria d'Université Côte d'Azur.Research ContextThe...
-
Phd Position F/m Fast Solvers for Studying Light
Il y a 5 mois
Sophia Antipolis, France Inria Temps pleinLe descriptif de l’offre ci-dessous est en Anglais_ **Type de contrat **:CDD**Niveau de diplôme exigé **:Bac + 5 ou équivalent**Autre diplôme apprécié **:Master in applied mathematics or scientific computing**Fonction **:Doctorant**Niveau d'expérience souhaité **:De 3 à 5 ansA propos du centre ou de la direction fonctionnelle The Inria centre at...
-
Phd Position, Scientific Machine Learning, Inria
Il y a 6 mois
Sophia Antipolis, France MathHire.org Temps pleinApplications are welcome at the Inria center of the University Cote d’Azur on the French Riviera situated in the vibrant technopole of Sophia-Antipolis, for a Master internship (4 to 6 months duration) followed by a three year fully funded PhD positon (to start immediately after or at commonly agreed date), in the context of a multi-partner collaborative...
