PhD Studentship: Applications of Machine Learning for Anomaly Detection in Financial Timeseries

University of Warwick

Funding: 4 years EPSRC funding for UK/EU students

Supervisors: Dr. Colm Connaughton (University of Warwick), Dr. Magnus Richardson (University of Warwick), and Dr. Marcus Ong (Spectra Analytics Ltd)

Supporting company: Spectra Analytics Ltd

Start date: October 2017

Project overview

Financial markets are highly complex systems with lots of non-linear interactions and feedback mechanisms. Given the importance of the financial system to society, as evidenced by the 2008 financial crisis, it is vital that we better understand its dynamics and can identify anomalies. These anomalies could range from illegal trading activities, such as insider trading, to fundamental changes in market dynamics, such as the impact of quantitative easing or currency manipulation by central banks. A greater understanding of these anomalies could be beneficial for market surveillance, economic regulation and the development of systematic trading strategies.

This PhD project will combine the domains of finance, machine learning, statistics and complexity science. It will develop new methods for anomaly detection based upon machine learning. The initial research will focus on detecting illegal trading and market manipulation both within and across financial assets. It has hitherto been very difficult to identify and prosecute offenders due to the sheer volume of traders and the complexity of their criminal activities. The research could then extend to examine other types of anomalies such as analysing the impact of news articles, tweets or company reports on financial asset returns.

The research builds upon proprietary machine learning algorithms currently being developed by Spectra Analytics and will use a combination of both real and synthetic trading data. It will use a variety of machine learning tools including classification/clustering analysis, regression analysis and natural language processing. Possible methods would include Gaussian Processes, Support Vector Machines, Convolutional Neural Networks, Sequential Monte Carlo and Probabilistic Graphical Models.

The PhD project is a collaboration between the EPSRC-MRC Centre for Doctoral Training in Mathematics for Real-World Systems (MathSys) and Spectra Analytics and is funded by the National Productivity Investment Fund.

Entry requirements

A 1st or 2:1 undergraduate degree and/or postgraduate Master’s qualification (MSc) in the mathematical sciences e.g. Mathematics, Computer Science, Statistics, Physics, etc.

Studentship funding

Eligible applicants will be paid a tax-free stipend of £14,553 pa (2017/18). UK applicants will be eligible for a full award paying home/EU tuition fees and maintenance. European Union applicants will be eligible for an award paying home/EU tuition fees only, except in exceptional circumstances, or where residency has been established for more than 3 years prior to the start of the course.


  • Make any initial enquiries by contacting the MathSys Director, Magnus Richardson (, in the first instance.
  • Complete the online application form - choose PhD in Mathematics of Systems (G1PG) and enter Dr. Magnus Richardson in the Finance section.
  • Upload a transcript from your current or previous study, a CV and a short cover letter explaining why you are interested in this project. Within 24 hours of submitting your online application you will receive a link to upload additional supporting documents.
  • When you submit your application, an email will automatically be sent to your referees requesting a reference. It will contain a secure link to upload a reference.

Deadline: Please submit your application by the 18 August 2017 for an interview shortly afterwards.

For more information please go to:

Share this PhD
  Share by Email   Print this job   More sharing options
We value your feedback on the quality of our adverts. If you have a comment to make about the overall quality of this advert, or its categorisation then please send us your feedback
Advert information

Type / Role:



Midlands of England