We have 3 PhD openings defined on the following themes:
1. AI-networking: decisions and clustering
The ability to accurately discover all hidden relations between items that share similarities is paramount to solving large optimization problems that pertain to artificial intelligence and networking. By embedding our recently developed clustering techniques into reinforcement learning problems, we will optimize several processes within KPN that relate to learning, decision taking, recommendation giving, and prediction making.
Background: probability theory, stochastic processes, network science, decision theory
Supervisors: Jaron Sanders and Piet Van Mieghem; Network Architectures and Services (NAS)
More information: nas.ewi.tudelft.nl, jaronsanders.nl
2. AI-networking: control and network science
Based on network state information (e.g. in routers), the network’s dynamic process is identified using system’s theory and new learning methods in order to control and manage the telecom network.
Background: network science, systems theory, telecommunications, stochastic processes
Supervisors: Bart De Schutter (3ME, DCSC) and Piet Van Mieghem (NAS)
More information: nas.ewi.tudelft.nl, www.dcsc.tudelft.nl
3. AI-networking: data and network science
Automatic recommendation of operation choices to network/system components (e.g. which content, service, or resource to allocate) when the system is subject to heterogeneous and dynamic user demand.
Background: network science, data science (recommender systems, machine learning, time series analysis), complex systems
Supervisors: Huijuan Wang and Alan Hanjalic; Multimedia Computing Group
For more information please visit the following link:
3 PhD Positions, TU Delft, Netherlands (2018)