Universidade de Vigo

Privacy metrics in federated learning. Ref. 2624-142770

OFFER REFERENCE 2624-142770

Open  19/11/2024
Deadline  25/11/2024

Indefinite contract

Work to be performed

The hired person will join the Signal Processing in Communications Group (GPSC, gpsc.uvigo.es) to work on the European project TRUMPET as part of the research team. They will co-lead the research on methods based on homomorphic encryption and differential privacy for machine learning, as well as the definition of privacy metrics in federated learning environments.

Functions to be performed:

  • Research in the design of cryptographic primitives to improve the privacy of federated learning.
  • Research on privacy metrics specific to federated learning environments.
  • Supervision of the implementation and testing of resulting methods for encryption, aggregation, and computation, and evaluation of their impact on data privacy.
  • Design of experiments to test the performance of the developed algorithms and protocols.
  • Research on possible optimizations for efficient implementations.
  • Preparation of scientific publications and project deliverables.
  • Representation of UVigo in project meetings

Required qualifications: PhD in Electrical/Telecommunications Engineering, Computer Science, Mathematics, or similar.

Gross monthly salary: 2.200,00 € (including the apportionments of extra payments)

Workday: 37.5 hours per week

Check all the info here:

CONCURSO DE MÉRITOS POR TRAMITACIÓN DE URXENCIA 26/24  PARA A CONTRATACIÓN DE PERSOAL INVESTIGADOR E DE APOIO Á INVESTIGACIÓN