The project VILAGIL aims to create an ecosystem capable of meeting the needs of the mobility for the Occitanie region, and more broadly to contribute to development innovations in the field of smart cities". As part of This project, the "Data and Mobility" action, designs and develops Mechanisms for automatic integration and federated access to mobility data. These data are provided by the project's non-academic partners such as that Toulouse Metropole, Sicoval, Tisséo Collectivité ; in some cases This data will be distributed on the different sites of the partners without possibility of integrating them centrally. The postdoc offer therefore targets particularly the issue of federated learning ("Federated") learning") to develop machine learning algorithms driven by this data.
Having regard to the high heterogeneity of the data collected in the VILAGIL project (data from sensors in the city surveillance images and videos, vehicle data autonomous, etc.), we want to offer a learning architecture Federated that considers data clients based on their nature (cross-device / cross-silo), and the nature of data distribution ( vertical/horizontal). This architecture should eventually ensure a Personalized learning for each client and fairness in relation to collaboration in learning.
We Let us consider developing this architecture by combining both a first Client clustering step and a Second stage of federated machine learning. Client clustering is often applied with the aim of identifying a subset of customers who would have similar learning models without disclosing their data. Approaches [1] and [2] propose methods applied in the context of non-IID data that calculate the similarities that would exist between Template settings to group customers. The objective of its approaches is to: Customize templates at local nodes taking into account heterogeneous distributions in the data. Drawing on recent work by [1], [2] The proposed architecture should be compared experimentally with other existing approaches to the literature [3] and demonstrating its interest.
The postdoc's mission will be:
From Develop a proof of concept to demonstrate architecture on identified data relevant to the project (real datasets of partners and/or datasets from the scientific literature of the domain).
From participate in meetings of the VILAGIL project at the level of the Action « Data & Mobility" and coordination meetings between the different actions.
From position its work and contributions in relation to the scientific literature in the field of federated learning, and publish contributions in World-leading international journals and conferences
Frame and environment
Laboratory: Institut de Recherche en Information de Toulouse, IRIT, CNRS/UMR5505
Scientific direction:
Imen Megdiche, GIS Team ([email protected])
André Péninou, Equipe SIG ([email protected])
Olivier Test, Team SIG ([email protected])
Place of Employment: IRIT Site Paul Sabatier, 118 Route de Narbonne, 31062 Toulouse, FRANCE
Duration: 1 year
Start-up: possible from January 2022 (FSD procedure required on the IRIT side)
Application file
We are looking for a Motivated candidate, force of proposals, with a solid background in mathematics applied and proven knowledge in AI.
Applications should be sent to ([email protected]) and ([email protected]) . The file must contain:
CV including scientific publications
Cover letter
Copy of last diploma
Pre-report and thesis defense report
Experience in federated learning would be an asset.
The LISTIC laboratory is recruiting an intern to work on federated learning.
The proposed activity will explore federated learning methods incorporating hierarchical partitioning processes. Considering a set of data sources, federated learning allows for local optimisation of a model and and to ensure a better generalisation by sharing information between the different local models. This approach is developing strongly, because it allows for better confidentiality and customisation by keeping the data and optimisation close to the sources.
However, many challenges remain to be overcome, in particular the construction of models capable of preserving the properties of the less represented sub-populations.
One avenue envisaged through this internship is the introduction of hierarchical clustering in federated learning. This approach could be studied in different application frameworks, in particular for inversion problems in geophysics, glaciers or even on data modelling from sensors related to intelligent housing or SmartCities.
A set of datasets and a model on the glaciology problem will already be available and will allow the internship to start. We are interested in estimating the depth of Swiss glaciers from different standard information such as slopes, velocities, temperatures.
However, depending on their location or the nature of the ground on which they slide, the training of the model could be adapted to obtain more accurate estimations. Given the large number of parameters capable of modifying the behaviour of a glacier and the difficulty of estimating their to estimate their impact, we propose to partition the glacier population automatically through federated learning in order to adapt the model parameters used to each group
Mickaël Bettinelli – [email protected]
Alexandre Benoit – [email protected]
Faiza Loukil – [email protected]
Realization of a software infrastructure for data collection and the analysis of the energy behaviour of an electric vehicle.
Primary duty station: LORIA
Contact : Vincent Chevrier (vincent DOT chevrier AT loria.fr)
The official job description (and therefore authentic) as well as the application (before December 13th) can be found at:
https://emploi.cnrs.fr/Offres/CDD/UMR7274-CARBON-001/Default.aspx
The CRIL (Lens Computer Science Research Lab - UMR 8188) Lab at University of Artois (Lens) is hiring a one-year postdoctoral researcher to work on the POSTCRYPTUM project, funded by the Agence Nationale de Recherche (ANR) and AID/DGA. This is a three-years project focused on algebraic cryptanalysis of public key cryptosystems, focused mainly on post-quantum schemes.
For more details on the project, please check here:
https://home.mis.u-picardie. fr/~ionica/postcryptum/Welcome.html
The ideal candidate should hold a Phd degree in Computer Science, Symbolic AI, propositional satisfiability (SAT) and beyond. Skills in one or several of the following topics will be appreciated:
SAT solving,
Problem encodings and reformulation,
Cryptography,
Pattern mining and machine learning.
The starting date is flexible, but preferably no later than February 2022.
For more information, please contact us ([email protected], [email protected], [email protected]).
IMT Atlantique (Campus Brest, France) is seeking candidates for a post-doctoral fellow position on Machine Learning.
This research is funded by Carnot TSN "Digital 2023" whose objective is to support groundbreaking exploratory projects or emerging from the state of the art combining digital technology and application fields.
The recruited person will have to invest mainly on the domain of anomaly detection in temporal networks, for specific contexts where explications are expected, i.e. where a decision has to be made regarding the detected anomalies and their properties. Part of the mission is the study and the synthesis of the recent works in both fields (anomaly detection in temporal graphs and anomaly explicability). The candidate will propose new methods, with a focus on unsupervised edge stream approaches.
The initial term of these positions is one year with the possibility for renewal upon performance review. The application deadline is December 4th, 2022, with the target start date January 3rd, 2023.
For details and instructions to apply, see:
https://institutminestelecom. recruitee.com/l/en/o/postdoctorante-ou-postdoctorant-en-detection-danomalies-dans-les-reseaux-temporels-cdd-12-mois
Candidates are encouraged to get in touch with us [email protected] questions.
A permanent position for a Senior Research Scientist is open in the Multidisciplinary Design Optimization (MDO) team of Applied Research and Technology (ART) at Collins Aerospace in Cork, Ireland.
Collins Aerospace, a unit of Raytheon Technologies Corp., is a leader in technologically advanced and intelligent solutions for the global aerospace and defence industry. ART is a global technology and open innovation resource within Collins Aerospace working on mission-critical projects that push the boundaries of what technology can do. We work on the cutting edge, redefining our industry with innovative partners, government and academia to research and advance transformative technologies that can create a safer, more connected and sustainable world. For additional information about ART, see:https://www.collinsaerospace. com/what-we-do/technology-and-innovation/applied-research-and-technology
This position focuses on combinatorial optimization, in particular constraint programming. Any other combinatorial optimization framework is a plus. For a detailed description of the position, see: https://careers.rtx.com/global/en/job/01536429/Senior-research-scientist-Combinatorial-optimization
The MDO team also has two other open positions for Senior Research Scientists: Aircraft systems design and optimization: https://careers.rtx.com/global/en/job/01581438/Senior-Research-Scientist-Aircraft-systems-design-and-optimization Structural optimization/System health monitoring: https://careers.rtx.com/global/en/job/01582675/Senior-Research-Scientist-Structural-optimization-System-health-monitoring
Please send your application before 20 December 23:59 to the address [email protected] (subject of the email:
post-doc application) by attaching the following documents:
CV containing a list of your publications, awards, etc.
Cover letter developing your experiences and contributions envisaged
Contact of one or more referees
Thesis reports Github software repository
As part of the DemoES project, the Catholic Institute of Lille is recruiting an XR research and development engineer over 18 months.
I send you the job description as an attachment, do not hesitate to send it to anyone interested.
To apply, please send your CV and a cover letter to Dr. Jalal Possik (jalal.possik@univ-catholille. (fr) before 9 December 2022.
we offer a full-time IGR position on a fixed-term contract for the year 2023 at the LISTIC laboratory (Univ. Savoie Mont-Blanc) on the Annecy-le-Vieux site.
Job details and application information here:
https://espaces-collaboratifs. grenet.fr/share/s/DFegtrxcRK-FuJhtUyoylg
We are recruiting a consultant in Constraint Programming with the following missions:
- the development of industrial applications of combinatorial optimization
- the development of global constraints for our internal solver
- participate in or bring a research project that fits into our themes
The position is for 12 months. Please share it with interested parties.
Arnaud Lallouet Principal Researcher, Head of Constraint Programming team
Huawei Technologies France Paris Research Center
20 quai du Point du Jour 92100 Boulogne-Billancourt France
Mobile: +33 661 776 220
E-mail: [email protected]
CANDIDATE
The candidate should have a PhD degree in signal/image processing and/or machine learning. Ideally, Python progaming language (and its learning modules : Keras, Pytorch, Tensorflow or Jax...) should be mastered. Lastly, knowledge in convex optimization is a plus.
TEAM
The candidate will be hosted in the computer science lab of the Institut de Recherche sur les Lois Fondamentales de l’Univers (IRFU) of CEA. She/he will be co-supervised with C.Kervazo from the Images group of Telecom Paris. This will provide the candidate with a rich environment in signal and image processing, machine learning and their applications.
Contacts : Jérôme Bobin [email protected]
Christophe Kervazo [email protected]
RESEARCH INTERNSHIP
Quantifying diversity of language phenomena: Case study of multiword expressions (LIFAT, Blois, France)
We propose a master internship position in Blois (France). Please send an email to apply, with a CV, a transcript of bachelor and master grades, and a few lines explaining your motivation to Arnaud Soulet [email protected] , as well as Agata Savary and Thomas Lavergne [email protected] .
Internship proposal description: https://selexini.lis-lab.fr/jobs/2022/11/26/internship
Application deadline: December 8, 2022 (or until filled)
MOTIVATION AND CONTEXT
Diversity of naturally occurring phenomena is a vital heritage to be preserved in the current progress- and optimization-driven globalization era. Diversity has been quantified in many domains: ecology, economy, information science, etc. but less so in Natural Language Processing (NLP). Recently, we have been addressing this aspect with respect to a particular linguistic phenomenon: the one of multiword expressions (MWEs).
MWEs, such as (FR) casser sa pipe ‘to die’ (literally to break one’s pipe) or (FR) sortir du lot 'to be better than others' (literally to quit the batch), are groups of words which exhibit unexpected properties (Baldwin & Kim, 2010; Constant et al. 2017). Most prominently, their meaning does not straightforwardly derive from the meanings of their components. Language resources dedicated to MWEs include MWE lexicons and MWE-annotated corpora (Savary et al., 2017), while a major computational task is to automatically identify MWEs in running text. The PARSEME network has been addressing the MWE identification task via a series of shared tasks on automatic identification of verbal MWEs (Ramisch et al. 2020). Our recent work (Lion-Bouton, 2021; Lion-Bouton et al. 2022) is explicitly dedicated to quantifying diversity in MWE language resources and MWE identification systems. We have adapted measures of variety (number of types in a system), balance (equity of items in various types) and disparity (differences between types), stemming notably from ecology and information theory (Morales 2021).
OBJECTIVE
The objective of this internship is to extend the formalisation of the diversity by benefiting from Good-Turing frequency estimation. Successfully used to estimate the biomass, Good-Turing frequency estimation is a statistical technique for estimating the probability of encountering an object of an unseen species, given a set of past observations of objects from different species (Good, 1953). Under this same principle, the idea would be to estimate the number of unseen MWEs from the MWEs observed in the corpus. Thus, it will be possible to correct the diversity measures to take the unseen MWEs into account and to evaluate the possible selection bias of the corpus.
IMT Atlantique/Lab-STICC/INRIA team Odyssey has several open 2-year post-doc positions on deep learning and computational imaging for ocean data assimilation and forecasting, more specifically on:
Deep emulators for marine plankton dynamics: - Deep learning and DA for ocean forecasting and reanalyses: https://cia-oceanix.github.io/downloads/postdoc_offer_anr_dream2022.pdf https://cia-oceanix.github.io/downloads/postdoc_offer_edito_model_lab2022.pdf
Learning-based reconstruction of upper ocean dynamics from underwater acoustics data: https://cia-oceanix.github.io/downloads/postdoc_OceaniXShom2022.pdf
These postdocs will explore how deep learning schemes provide new means to revisit and advance inverse problems in ocean modeling and monitoring. More details in the above links and https://cia-oceanix.github.io. Targeted contributions cover demonstrations from idealized/simulated systems to real satellite/in situ observation datasets. We welcome both applicants with a background in applied math./geoscience and interest in deep learning as well as applicants with a background in data science and interest in ocean science.
Deadline for applications: the review of applications will begin immediately and continue until the positions are filled.