Description
Date depot: 11 mars 2020
Titre: Analysis and Optimization of Collaborative Machine Learning
Directeur de thèse:
Maurizio FILIPPONE (Eurecom)
Domaine scientifique: Sciences et technologies de l'information et de la communication
Thématique CNRS : Non defini
Resumé:
With their evolution towards 5G and beyond, wireless communication networks are entering an era of massive connectivity, massive data, and extreme service demands.
A promising approach to successfully handle such a magnitude of complexity and data volume is to develop new network management and optimization tools based on machine learning.
This project will study and develop novel solutions in this domain, relying in particular on Bayesian Machine Learning techniques, which make it possible to reason about the level of uncertainty in predictions, and collaborative neural networks, which allow one to model interactions in communication systems.
In particular, the objective of the Ph.D. will focus on (i) improving the training efficiency of collaborative neural networks and study their properties; (ii) study novel ways to carry out model selection by employing Bayesian techniques; (iii) combine advancements in machine learning and computer vision with Bayesian methods in order to tackle novel applications in communication systems.
It is expected that thanks to the application in communication systems, there is going to be the need to impact the field of Bayesian machine learning, in a cross-fertilization spirit.
Doctorant.e: Gogolashvili Davit