Research Interests

  • Machine Learning: Making computers do the work. Enabling systems to learn, generalise and reason is challenging but through the successful application of artificial intelligence we can not only generate previously untapped value from data but also automate complex operational processes. My particular area of interest is the application of “online” training of deep neural networks for prediction in non-stationary environments.
  • Complex Adaptive Systems: I have a general interest in evolutionary simulation modelling, self-organising and adaptation systems. After spending a year in the Master’s programme at the Institute for Complex Systems Simulation I lead on a project that used artificial evolution and agent-based modelling to generate autonomous adaptive algorithmic trading strategies and market structures that improved stability and resilience of financial markets.
  • Automated Trading Systems: A large portion of my PhD research involves the development and study of automated algorithmic trading systems. Such systems: generate signals indicating whether to buy, sell or hold an asset (trading signals); efficiently manage their own risk; and execute their trading decisions while minimising market impact. These quantitative trading strategies attempt to gain an edge by finding trends or indicators from historical data. We do this using time series analysis, artificial intelligence and machine learning to isolate useful information from the mass of available data.
  • Semi-synthetic Agent-based Modelling of Order Books: I am currently working on a project that uses level 2 order book data to drive agent-based models for the purpose of evaluating trading strategies. Such models allow us to generate alternative scenarios from historical data that maintain all of the statistical properties of markets but also allow us to generate a variety of alternative outcomes.