Skip to content

Stochastic Modelling and Operations Research

The Stochastic Modelling and Operations Research Group (SMORG) fuses theory, computation, and data, to address exciting and important problems in networks and social science, and across ecology, epidemiology and evolution.Our award-winning staff currently supervise over 15 research students,creating a stimulating group in which to produce world-leading research.

SMORG provides a vibrant and national-leading environment for research. Our award-winning staff currently supervise over 15 research students,creating a stimulating group in which to produce world-leading research

SMORG’s staff includes medallists of the Australian Academy of Science and the Australian Mathematical Society, Australian Research Council Fellowship recipients, and are key members of the Australian Research Council Centre of Excellence for Mathematical and Statistical Frontiers (ACEMS)

Our recent PhD graduates have been employed around the world, including in academia – for example, at the University of Cambridge – in industry – for example, at AT T, Microsoft and Quantium – as well as in defence.

Researcher and Interests
Nigel Bean Matrix-analytic methods in stochastic modelling, Markov chains, stochastic modelling; network modelling; operations research; mathematical modelling
Andrew Black Stochastic epidemic modelling, evolutionary dynamics, complex systems and stochastic modelling
Robert Cope Effective parametrization and forecasting in stochastic epidemic models based on a variety of data sources and at different spatial scales; Ecology, including developing models for the transport of invasive species.
David Green Matrix Analytic Methods, Markov chains, Hidden Markov models, Optimisation, Stochastic Dynamic Programming and Simulation Modelling
Lewis Mitchell Social networks and social media analysis; human dynamics; natural language processes; ensemble data assimilation and prediction
Giang Nguyen Matrix-analytic methods, stochastic fluid processes, Brownian motion
Joshua Ross Modelling in Ecology, Epidemiology and Evolution; Stochastic modelling; Bayesian computational statistics; and Operations Research
Matthew Roughan Network tomography and traffic matrix modelling, routing robustness, compressive sensing, privacy preserving data mining and management
Andrew Smith Stochastic processes and dynamical systems applied to populations and more specifically metapopulations
Silvio Tarca Undertaking research on modelling and control of a future electricity grid with a high penetration of power electronic converter connected devices, distributed and utility scale, including renewable energy generators, battery energy storage systems, electric vehicles and electrical loads

Research Areas

Computational Probability




Ecology, Epidemiology and Evolution


Possible PhD Projects

  • Predictability in social networks - Dr Lewis Mitchell

    The Internet has the potential to make the world increasingly interconnected, with social media providing both new ways for people to interact, and new datasets with which to study human dynamics. But how much information exists in such social networks, and to what extent can human behaviour be predicted from observing online interactions?

    Projects will look at topics like:

    • entropy estimation in social networks
    • link prediction
    • sentiment analysis
    • dynamics of information flow over networks
  • Health system modelling - Professor Nigel Bean

    With colleagues across the country, I am involved in research work modelling the health system. The health system is an incredibly complicated network of queues (queues at the Emergency Department (ED) or at the GP; people in ED then queue to get into a ward; people in wards queue to get into rehabilitation facilities, or old-aged homes). We have had in-depth involvements with a number of the major South Australian hospitals. That specific commercial work is not directly suitable to research students, however, we find that really interesting side projects arise from such work.

    A specific example arises from work we did modelling the RAH Intensive Care Unit. Essentially all queueing theory assumes that the arrival process and the service times (called length of stay in a hospital system) are statistically independent. For some reason, one of the first things we did with the ICU data was to check this fundamental fact (the process for doing that is an interesting piece of research in its own right). As we were doing a commercial project, we then had to find a quick work-around to enable us to still provide them with the necessary answers; but the question remains: why is this the case and can we use the data to identify the mechanism and the extent of this effect in order to build better models that directly account for this dependence?

    Current progress suggests (not unreasonably) that if the ICU is fuller, then they are more likely to block entry to people in ED (or on a ward) - remember I said that the health system was a network of queues? - and thus reduce the arrival rate. However, this cannot be the entire answer as we also tested if the dependence could be explained solely through the occupancy status of the ICU, and it couldn’t. This is by no means the only project available in this area, but it hopefully gives you a flavour of the sort of project we might be able to work on. We can also do more directly applicable projects investigating specific questions of interest in specific health-system settings.

    Health System Modelling

  • Multi-level models of evolutionary transitions - Dr Andrew Black
    Evolutionary Transitions

    Evolutionary transitions are processes responsible for creating new levels of biological organisation. For example, multicellular organisms are composed of single cells and chromosomes are collections of individual genes. One view on these transitions is that ecology plays an important role in creating the conditions that then lead to the emergence of higher-level entities.

    This project will use stochastic models to look at how different ecologies can create multi-level Darwinian populations and examine how their evolutionary properties are linked to the ecological properties. This work has potential applications for creating artificial multi-level populations either to evolve new types of multicellular organisms or to select for cellular traits that would otherwise require genetic engineering to produce.

  • Simulation methods for epidemic model inference - Dr Andrew Black and A/Prof Joshua Ross

    Inference for epidemic models is a rapidly advancing field. Many state-of-the-art methods, such as particle MCMC, rely on repeatedly simulating realisations of the underlying model. Producing these can be computationally expensive and can restrict the models we can use. This project will use techniques such as importance sampling to develop better methods for simulating from these models. The methods developed will be used for inference on novel historical datasets such as from the 1918 Spanish ‘flu pandemic, or more recent outbreaks such as Ebola.

    Evolutionary Transitions

  • Stochastic Processes- Dr Giang Nguyen

    I am happy to supervise students in projects related to Markov chains, Brownian motions, diffusion processes, etc., and their applications.

    ">Stochastic Processes

  • The Mathematics of Energy Systems- Dr Giang Nguyen

    I am happy to supervise students in projects related to Markov chains, Brownian motions, diffusion processes, etc., and their applications.

School of Mathematical Sciences
Level 6, Ingkarni Wardli

North Terrace Campus
See location on map


General email
Head of School email
Telephone: +61 8 8313 5407
Facsimile: +61 8 8313 3696

Find Us Online

School of Maths on Facebook    School of Maths Blog