PhD Position in Machine Learning

Project Description

PhD Area:

- Machine Learning

- Software Defined Networks

- Computationally Efficient Learning

- Dependable Networks and Customised Networks

- Data-driven Optimization and Managemen

System dynamics create challenges in the ways that machine learning is applied to the metrics drawn from communications networks. To engineer future network systems that offer high service dependability in the context of Software Defined Networks, three things must be done:

(1) the performance of the underlying network switches must be mapped to the services delivered;

(2) awareness of the state of the factors influencing such dependability, including current device load, network congestion and throughput needs to be accounted for; and finally,

(3) learning must be completed fast, so that insights can be acted upon.

In this PhD, provision of advance notice of future events via predictive monitoring will be achieved quickly, by developing algorithms that optimize the way in which subsets of historical data are selected for learning and optimising the efficiency of the learning algorithms that operate on this data. The research theme of this PhD project falls into the following areas: ``Dependable Networks and Customised Networks'' and ``Data-driven Optimization and Management'' of programmable networks. It builds on recent work in the PI's team in the area of fast Quality of Delivery prediction and learning algorithms for dynamic networks.

The successful candidate will gain practical experience developing research testbeds and implementations of fast learning algorithms. The successful candidate will carry out research as part of the SFI CONNECT research centre, https://connectcentre.ie/.CONNECT is Science Foundation Ireland’s world leading research centre in Future Networks and Communications.

If you are interested in submitting an application for this project, please email your CV and 2-page letter of motivation to [email protected]