PhD Position F/M: Robust Reinforcement Learning and Privacy in Sequential Decision Making

il y a 4 semaines


Villeneuved'Ascq, Hauts-de-France INRIA Temps plein

Job Context and Requirements

We are seeking a highly motivated PhD candidate to join our team at INRIA and work on a project focused on robust reinforcement learning and privacy in sequential decision making. The candidate will be supported by the PEPR project FOUNDRY and supervised by Debabrota and Emilie.

As RL algorithms are being deployed in real-life scenarios, questions of responsible deployment arise, such as robustness to noise and perturbation in the feedback from the environment, and privacy if users are involved in the environment yielding data.

Our previous works have shown that for structure-less and linear settings of multi-armed bandits and active testing, imposing privacy yields two regimes of performance. For the regime used in practice, privacy can be preserved without loss of utility. However, our existing approach is not directly applicable to more practically appealing settings of RL, like MDPs or bandits with side-information. Thus, we want to study whether the cost of privacy in contextual bandits and MDPs, and also design optimal, computationally efficient algorithms.

Similarly, we have studied the impact of unbounded corruption in feedback and safety constraints in stochastic multi-armed bandits and active testing. We want to understand how these factors impact more structured RL problems and how we can design optimal algorithms in these settings.

The project is expected to simulate existing and new collaborations with researchers and groups working on privacy-preserving machine learning, robustness, adaptive testing, and reinforcement learning. In the future, the candidate will be encouraged to collaborate internationally and be part of the INRIA community.

Job Mission

This position is dedicated to completing a PhD thesis. French rules emphasize that the PhD should be completed within 3 full years of studies. It is also possible to teach up to a reasonable amount of time per year (approximately 30 hours/year).

Main Activities

The candidate will be responsible for all research activities, including bibliographical search, proposing original ideas related to the topic of the PhD, developing them, presenting the work in the INRIA seminar, workshops, and conferences. The candidate should aim to publish research results in premier conferences and journals of our field of research (e.g. ICML, NeurIPS, COLT, IJCAI, AAAI, JMLR). Since the work involves and impacts responsible AI in general, the successful candidate should collaborate in writing scientific articles for a larger audience.

Required Skills

  • A strong background in mathematics/statistics
  • A good knowledge of machine learning, statistics, and algorithms
  • Broad interest in differential privacy and robustness
  • Knowledge of programming languages such as Python and C/C++
  • Some experience with implementation and experimentation (a plus)
  • A good command of English

Please follow the instructions given in the application guidelines to set up your application file. The application should include a CV, an application letter, (two or more) recommendation letters, and school transcripts. It is recommended that the candidate contacts Debabrota and Emilie while preparing the application.

Benefits

  • Subsidized meals
  • 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
  • Access to vocational training
  • Social security coverage

Salary

1st and 2nd year: 2100 € (gross monthly salary)

3rd year: 2190 € (gross monthly salary)



  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Context and Key Features of the PositionThe selected candidate will be supported by the PEPR project FOUNDRY and supervised by Debabrota and Emilie, affiliated with the INRIA.As RL algorithms are increasingly deployed in real-life scenarios, questions arise regarding responsible deployment, such as robustness to noise and perturbations in the feedback from...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Postdoctoral Researcher PositionINRIA is seeking a highly motivated postdoctoral researcher to work on understanding the impact of different types of constraints and structures on the performance and design of bandit and reinforcement learning algorithms.Research ObjectivesThe postdoctoral researcher will investigate the fundamental question of reinforcement...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Context and advantages of the positionThe successful candidate will be supported by the PEPR project FOUNDRY and supervised by Debabrota and Emilie, affiliated with the INRIA. The project aims to study the impact of privacy in contextual bandits and MDPs, and to design optimal, computationally efficient algorithms.Main activitiesThe candidate will be...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Context and Benefits of the PositionThe postdoctoral researcher will be part of the INRIA team and will work under the supervision of Debabrota. The researcher will investigate the impact of different types of constraints and structures on the performance and design of bandit and RL algorithms.The project aims to simulate existing and new collaborations with...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Research ContextThe selected candidate will be part of the research team at INRIA, working under the supervision of Debabrota. The objective of this postdoctoral position is to investigate the impact of different types of constraints and structures on the performance and design of bandit and reinforcement learning algorithms. This research aims to simulate...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job Context and RequirementsThis PhD position is part of a project that aims to develop privacy-preserving federated learning algorithms. The successful candidate will work in the MAGNET team at INRIA, collaborating with European project partners.With the increasing use of AI, there is a growing concern about potential risks and abuses. Research directions...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Research Position in Secure Machine LearningThis PhD student position is part of the MAGNET team at INRIA, focusing on research and development of privacy-preserving federated learning algorithms. The recruited PhD student will collaborate with colleagues in the MAGNET team and the project consortia to contribute to the TRUMPET, FLUTE, and REDEEM...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Research Context and ObjectivesThis PhD position is part of a project that aims to develop privacy-preserving federated learning algorithms for applications in oncology. The successful candidate will contribute to the research and development of these algorithms in collaboration with the MAGNET team and project partners.Key ResponsibilitiesContribute to the...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Research Opportunity in Privacy-Preserving Federated LearningThis PhD student position will be supported by the project, allowing for close collaboration with national and international partners. The successful candidate will contribute to research and prototyping algorithms for secure, privacy-preserving federated learning in settings with potentially...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job ContextThe INRIA team is seeking a postdoctoral researcher to contribute to the development of privacy-preserving federated learning algorithms and their applications in medical domains.The successful candidate will work closely with the MAGNET team and European project partners to advance the field of privacy-preserving machine learning.Research...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job Context and RequirementsThis postdoctoral position will be supported by the project and will be part of the MAGNET team in Lille, collaborating with several European project partners.With the increasing concern about potential risks and abuses of AI techniques, there is a growing interest in research directions such as privacy-preserving machine...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    About the RoleThis position will be part of the MAGNET team in Lille, collaborating with European project partners. The successful candidate will contribute to researching and prototyping algorithms for secure, privacy-preserving federated learning in settings with potentially malicious participants.Key ResponsibilitiesDeveloping algorithms, such as...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Context and Advantages of the PositionThis postdoctoral position will be supported by the project, allowing for collaboration with several European project partners. The MAGNET team, involved in the TRUMPET, FLUTE, and REDEEM projects, seeks team members who can contribute to these projects in close collaboration with other team members and national &...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Context and Highlights of the PositionThis position will be supported by the INRIA project. As part of the MAGNET team in Lille, the engineer will collaborate with several European project partners.Given the growing concern about potential risks and abuses of AI, there is a significant interest in research directions such as privacy-preserving machine...

  • Postdoctoral Research Fellow

    il y a 3 semaines


    Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Research Opportunity We are seeking a highly motivated postdoctoral researcher to join the MAGNET team at INRIA. The successful candidate will contribute to the development of privacy-preserving federated learning algorithms for medical applications.Context and ObjectivesThis postdoctoral position will be supported by the project, which aims to research and...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job Opportunity: Scientific Programmer in Federated Learning for Oncology ApplicationsThis position is part of the MAGNET team at INRIA, collaborating with European project partners. The team is involved in research projects such as TRUMPET, FLUTE, and REDEEM, focusing on privacy-preserving federated learning in settings with potentially malicious...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job Opportunity: Scientific Programmer in Federated Learning for Oncology ApplicationsThis position is part of the MAGNET team at INRIA, collaborating with European project partners. The team is involved in projects such as TRUMPET, FLUTE, and REDEEM, focusing on privacy-preserving federated learning in settings with potentially malicious participants.Key...


  • Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Job Context and RequirementsThis position will be supported by the project and will collaborate with several European project partners. The recruited engineer will contribute to one or more of the TRUMPET, FLUTE, and REDEEM projects, which aim at researching and prototyping algorithms for secure, privacy-preserving federated learning in settings with...

  • Scientific Researcher

    il y a 2 semaines


    Villeneuve-d'Ascq, Hauts-de-France INRIA Temps plein

    Researcher Position - Federated Learning in Oncology This researcher position will be supported by the project, with collaboration between the MAGNET team in Lille and European project partners. Research on privacy-preserving machine learning is gaining importance due to concerns about potential risks and abuses. Techniques such as differential privacy and...


  • Villeneuve-d'Ascq, Hauts-de-France Inria Temps plein

    Job Title: Engineer ContractorWe are seeking a highly skilled Engineer Contractor to join our team at Inria, a leading research institute in the field of digital sciences and technologies. The successful candidate will be responsible for consolidating and extending an existing open-source library for decentralized and privacy-preserving machine learning.Key...