Learning the Structures of Models of Computer System Performance

Institution:

University of Edinburgh

Dept/School/Faculty:

School of Informatics

PhD Supervisor:

Dr C Sutton

Application Deadline:

Applications accepted all year round

Funding Availability:

Competition Funded PhD Project (Students Worldwide)

Modern computer systems have become more complex than ever before, with distributed systems becoming a mainstream computing tool. Low latency is a crucial design goal for these systems, because users will not adopt an interactive Web service that is slow. Understanding the performance of a distributed system is extremely difficult because of the many intercations between components.

In this project, we will address this problem by attempting to learn the structure of models to describe the performance of these systems. The goal of this projects is to automatically determine the structure of models to describe the performance of warehouse-scale and cloud applications. Possible structure may include networks of nonparametric regression models, networks of queues, or more complex performance models such as stochastic process algebras. The idea is that the learning structure will be useful for visualization, i.e., that it will provide a compact, interpretable description of the system's performance, so that performance bugs in the system will be visually apparent as bottlenecks in the learned queueing network. Essentially, the learned model will serve as a summary of the large amount of performance data used to generate it. Structure learning is a notoriously complex problem in machine learning, so this new application may serve as a challenge problem for this area.

This is an opportunity to join a world-leading research group in machine learning. The machine learning group in the Institute for Adaptive and Neural Computation is made up of 6 academic staff: Chris Bishop, Chris Williams, Amos Storkey, Charles Sutton, Guido Sanguinetti, Iain Murray. The group conducts research into the development of novel probabilistic machine learning methods, and the application of these novel methods to cutting edge scientific/technological problems.

The project is suitable for a student with a good background in mathematics and in computer science, ideally a first class degree in computer science, mathematics, physics, or engineering. Some knowledge of machine learning is highly desirable, as well as some programming ability. Informal enquiries should be addressed to Dr Charles Sutton at csutton@inf.ed.ac.uk. We recommend that students make informal contact before making a formal application, ideally as soon as possible: the more time between informal contact and the application deadline the better. Application deadlines match the standard Informatics deadlines at http://www.ed.ac.uk/schools-departments/informatics/postgraduate/apply/keydatesresearchappns.

We advise students to apply before the 16 December 2011 deadline.

Formal application must be through the School's normal PhD application process: http://www.ed.ac.uk/schools-departments/informatics/postgraduate/apply Select the Informatics: Institute for Adaptive and Neural Computation research area. 

Funding Notes:

The School offers a variety of scholarships, see http://www.ed.ac.uk/schools-departments/informatics/postgraduate/fees.