The Wynyard Group specialises in powerful software to help protect companies and countries from threat, crime and corruption. We are offering a scholarship to a suitable candidate to pursue a PhD in Statistics, in the area of Bayesian predictive analytics. To qualify, you must be eligible to enrol as a PhD candidate in the Department of Mathematics and Statistics at the University of Canterbury in Christchurch, New Zealand. To check your eligibility, please refer to the enrolment information in: http://www.canterbury.ac.nz/postgrad/phd_students/enrolment_fees.shtml
If you are offered the scholarship, you will be expected to enrol as a full-time PhD candidate in the Department of Mathematics and Statistics, University of Canterbury, starting in February 2014. You will be jointly supervised by a Wynyard Group staff member and an academic from the University. You will be required to spend about half of your time at Wynyard and the other half at the University, both located in Christchurch.
The scholarship is worth NZ$30,000 per annum and will be tenable for 3 years. It will be renewed annually subject to satisfactory progress in your research. The scholarship does not cover any other expense. You will be responsible for all other expenses, including the University’s enrolment fees (currently about NZ$6,500 per annum).
Project title: Bayesian anomaly detection in noisy dynamic graphs.
Project description:
This project seeks to develop, implement and test Bayesian models for detecting anomalies in network data that are time-varying and noisy. The data will be represented by random attributed graphs. An attributed graph is a collection of vertices and edges augmented by vertex and edge attributes.
Some examples of applications producing data that are representable as attributed graphs include crime detection and prevention, infectious disease transmission, targeted marketing, and protein interaction. In such applications, anomalies within a graph or appearing over time are often of interest. In crime detection and prevention, for example, anomalies may indicate criminal activity or impending criminal activity. In disease transmission, anomalies may signal the source of an outbreak or the existence of naturally-occurring resistance. The data in these applications are often dynamic and noisy, which further complicate the detection of anomalies. A computational Bayesian approach will be taken, involving semiparametric and nonparametric hierarchical models. Efficient implementation of these models will utilize advanced Markov chain Monte Carlo and variational methods.
Candidate profile:
You must have a Bachelor's degree with Honours (equivalent to a 4-year Bachelor's degree in the US) or a Master's degree, in Statistics, Mathematics or Computer Science, with outstanding grades. You must be independent, innovative, self-motivated, disciplined, perserverant and love doing research. You must have good oral and written communication skills, be fluent in English and must satisfy the University's English language requirements.
You must have excellent statistical, mathematical and computational skills, including substantial experience in programming (preferably in Matlab/Octave, R and C++/C#). You must have knowledge of Bayesian statistics, graph theory and grahical models, at least up to intermediate-undergraduate level.
Application:
To apply, please send (1) a detailed CV, (2) detailed transcripts of all university results, and (3) a write-up describing your research interests and aspirations and why you think you are a match for this project. Please send these as attachments in an email with subject heading, "Wynyard PhD Scholarship application", to:
Dr. Dominic Lee, Email: This email address is being protected from spambots. You need JavaScript enabled to view it.
For queries about enrolment in the PhD programme at the University of Canterbury, please send email to: https://wynyardgroup.com/
University of Canterbury: http://www.canterbury.ac.nz/
Department of Mathematics and Statistics: http://www.math.canterbury.ac.nz/
Please refer to http://www.canterbury.ac.nz/emaildisclaimer for more information.
