Offre en lien avec l’Action/le Réseau : – — –/– — –
Laboratoire/Entreprise : I3S
Durée : 12 months
Contact : firstname.lastname@example.org
Date limite de publication : 2023-05-31
This postdoc is part of the ANR ARTIC project (ARTificial Intelligence-based Cloud network control, cf. http://www.i3s.unice.fr/~raparicio/project/artic/), of which Ramon APARICIO PARDO is the principal investigator. This project will provide the candidate with the funds and resources necessary for their activities (participation in scientific events, equipment, computer, access to computing platforms, etc.)
The postdoc will take place in the I3S laboratory, a joint public research laboratory resulting from the collaboration of the CNRS, Univ. Cote d´Azur and INRIA. The postdoc will work with experts in optimization, machine learning and telecommunications networks from the I3S and INRIA.
Nowadays, cloud IP traffic has become the most part of Internet traffic. A traffic that complexifies with an increasing devices diversity and traffic dynamicity . The combination of Machine Learning (ML) and Artificial Intelligence (AI) with Network Softwarization (SDN/NFV). has been proposed in the so-called Knowledge Defined Networking (KDN) to give rise to a “Cognitive Cloud:” an AI-enabled Cloud-Edge framework. This “Cognitive Cloud” will allow automatically adapting to the growing complexity and variability of Internet traffic by (i) (re)learning Cloud-Edge network control policies from data; and (ii) applying these control policies onto a (re)configurable Cloud-Edge network.
In many cases, the management of this (re)configurable Cloud-Edge network constitutes a challenging stochastic control problem, that can be modelled as Markov Decision Process (MDP) and solved under the Reinforcement Learning (RL) framework (a form of Machine Learning). Novel RL approaches could allow us to find more efficient management decisions (i.e., the control policy) to operate this Cloud-Edge network.
Then, in this postdoc, we aim to apply ML (as RL, but not uniquely) framework to the management of Cloud-Edge networks.
Profil du candidat :
Diploma required: PhD degree less than 1-2 years old in computer science / mathematics / telecommunications, with solid experience in machine learning (artificial neural networks) and computer and telecommunication networks
Formation et compétences requises :
– Python 3.5 language, Python frameworks (like PyCharm, Jupiter Notebook, Spyder, Conda)
– Deep Learning Libraries (like TensorFlow, Keras)
– Networks and system (Unix, typically)
– Machine learning and data science (namely neural network theory)
– Classical optimization theory (convex optimization, combinatorial optimization)
– Computer network control plane (algorithms and protocols)
Adresse d’emploi :
To obtain more information, send an e-mail to email@example.com
Supporting documents for the application
• 1. Curriculum vitae, with list of publications
• 2. Cover letter
• 3. PhD diploma
• 4. PhD dissertation
• 5. Thesis (pre-)defense reports (if available)
• 6. At least two letters of recommendation and a list of three references to contact.