Post-Doctoral Research Visit F/M Risk Measures in Task-Oriented Communications

The Inria research centre in Lyon is the 9th Inria research centre, formally created in January 2022.  It brings together approximately 300 people in 16 research teams and research support services.

Its staff are distributed at this stage on 2 campuses: in Villeurbanne La Doua (Centre / INSA Lyon / UCBL) on the one hand, and Lyon Gerland  (ENS de Lyon) on the other.

The Lyon centre is active in the fields of software, distributed and high-performance computing, embedded systems, quantum computing and privacy in the digital world, but also in digital health and computational biology.

This postdoctoral research position will be carried out in Inria Lyon, funded by the PEPR Networks of the Future programme. The candidate will be hosted within the MARACAS team in the CITI Laboratory, working primarily with Dr. Malcolm Egan. As the PEPR Networks of the Future programme is a national French collaboration, opportunities will be available for interactions with other researchers working in foundations and applications via seminars, workshops, and short-term visits.

The focus of the PEPR Networks of the Future programme is to develop fundamental and applied research towards future communication networks. One aspect of this programme is goal-oriented and semantic communications, where communications is tailored for a specific task (e.g., training machine learning models or process control). In addition to considering task-dependent constraints (e.g., latency or reliability), a key aspect of goal-oriented communications is selecting relevant data to communicate.

In the work to be carried out in this postdoctoral position, the focus is on coding schemes (both compression and channel) with guarantees on the performance of a specific task. In particular, unlike for traditional coding schemes, we will focus on risk constraints, which account for the impact of a large distortion in the received signal on the performance of a task. Our focus will be on tasks modeled via data-dependent optimization, which arises in machine learning, statistical inference, and resource allocation.

The goal of this postdoctoral position will be to design and analyze coding schemes using tools from information theory subject to risk constraints where the distortion measure is defined in terms of optimality loss in data-dependent optimization. The performance of algorithms to construct (joint) source and channel codes will be then evaluated in comparison with the theoretical analysis.

The candidate will carry out research on federated learning algorithms in collaboration with members of the MARACAS Inria project-team. This includes participation in local seminars as well as in summer schools and international conferences.

Main activities:

• Information-theoretic analysis of risk-constrained source and channel coding.
• Development of algorithms to construct source and channel codes with risk-constraints.
• Simulation-based comparisons of information-theoretic bounds and proposed algorithms.

• Subsidized meal
• 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 (90 days / year) and flexible organization of working hours
• Professional equipment available (videoconferencing, loan of computer equipment, etc.)
• Social, cultural and sports events and activities
• Access to vocational training
• Complementary health insurance under conditions

2788€ gross salary / month