Engineer F/M - Scientific programmer in federate learning with applications in neuroscience

Inria; Villeneuve-d'Ascq; Hybrid; EUR 45,000 - 65,000; Full time

Benefits offered by this job

Public transport refund

Annual leave: 7 weeks + 10 RTT days

Teleworking options

Equipment provided

Cultural and sports events

Job summary

Inria Lille accueille le poste d'Ingénieur logiciel scientifique spécialisé en apprentissage fédéré appliqué à la neuroscience. Le candidat travaillera sur le projet Loop, intégrant des techniques fédérées pour des données FRMI/EEG et collaborera à l'extension de DecLearn dans un cadre décentralisé.

Le rôle exige une maîtrise des algorithmes distribués et des compétences en développement logiciel (Python/Rust).

Qualifications

• Forte compréhension des algorithmes distribués.
• Compétences en conception et développement logiciel (Python et Rust possible).

Responsibilities

• Rejoindre l'équipe qui écrit le pipeline de préparation des données.
• Adapter et mettre en œuvre des algorithmes fédérés adaptés au problème d'apprentissage.
• Étendre DecLearn vers un cadre entièrement décentralisé.
• Aborder les questions de confidentialité dans les expériences d'apprentissage fédéré.
• Mettre en œuvre des approches pour le calcul neuromorphique.
• Développer le code et écrire des tests.
• Lancer les expériences et rendre compte des activités au sein du consortium.

Skills

Distributed algorithms

Python

Rust

Tools

Python

Rust

Job description

Engineer F/M - Scientific programmer in federate learning with applications in neuroscience

Created in 2008, the Inria center at the University of Lille employs 360 people, including 305 scientists in 16 research teams. Recognized for its strong involvement in the socio-economic development of the Hauts-De-France region, the Inria center at the University of Lille maintains a close relationship with large companies and SMEs. By fostering synergies between researchers and industry, Inria contributes to the transfer of skills and expertise in the field of digital technologies, and provides access to the best of European and international research for the benefit of innovation and businesses, particularly in the region.

For over 10 years, the Inria center at the University of Lille has been at the heart of Lille's university and scientific ecosystem, as well as at the heart of Frenchtech, with a technology showroom based on avenue de Bretagne in Lille, on the EuraTechnologies site of economic excellence dedicated to information and communication technologies (ICT).

The person will join the cross disciplinary project Loop (closed Loop neurOtechnologies: from sensOrs to aPplications). The project will implement an interdisciplinary strategy applied to the development of a mechanism-based therapy for drug-resistant hallucinations in patients with schizophrenia. The project will be structured around three main research pillars: 1) the algorithmic level of closed-loop systems (online signal processing, machine learning); 2) the design of a complete and non-invasive solution to address the question of refractory hallucinations, and study how closed-loop interaction schemes can be beneficial for highly impaired psychiatric patients; 3) the design of the sensors themselves and translational research with animal models, in order to enhance the quality of the recordings and pave the way for future invasive miniaturized solutions.

In this context, the objective is to apply federated and decentralized learning techniques to learn predictive models.

The objective is to adapt the declearn API to the specific case of FRMI and EEG data, time series and eventually to address the problem of federated learning in the neuromorphic paradigm.

The work plan is the following

• Join the team that writes the data preparation pipeline. The challenges are the scalability with respect to data size, data dimensionality, and time dependent data. The declearn readers and data representation choices will be monitored and adapted to the new choices and constraints.
• Adapt and implement federated algorithms tailored to the learning problem specifications. A specific attention will be on the scalability and the efficiency of the communications.
• Extend DecLearn to the fully decentralized setting.
• Address the problem of privacy.
• Implement specific approaches for neuromorphic computation.
• Develop code.
• Write tests.
• Launch experiments.
• Report activity in the consortium.

Technical skills and level required :

• a strong understanding of distributed algorithms
• software design and development skills (relevant code may include Python and Rust)

Languages :

Relational skills :

• smoothly working in a team in a reseach environment
• effective communication and collaboration
• eager to learn in an academic setting

Avantages

• Partial reimbursement of public transport costs
• Leave: 7 weeks of annual leave + 10 extra days off due to RTT (statutory reduction in working hours) + possibility of exceptional leave (sick children, moving home, etc.)
• Possibility of teleworking and flexible organization of working hours
• Professional equipment available (videoconferencing, loan of computer equipment, etc.)
• Social, cultural and sports events and activities