Skip to main content
  • Undergraduate Studies
  • Graduate Studies
  • Diploma
  • Entry Scholarships
  • Student Resources
  • Academic Advising
  • Learning Centre
  • Minors
  • Undergraduate Research Opportunities
  • Funding & Scholarships
  • Course Outlines
  • Awards Night
  • About Us
  • Contact Us
  • Our People
  • News
  • Events
  • Math & Stats Club
  • Supermath Day
  • Kangaroo Contest
  • Math Cards
  • Biomathematics Seminar Series
  • Data Science Speaker Series
  • Department Colloquium Series

    Alex Shestopaloff

    Headshot of Alexander Shestopaloff

    Assistant Professor

    College of Computational, Mathematical and Physical Sciences, Department of Mathematics & Statistics

    ashestop@uoguelph.ca
    Google Scholar
    Available positions for grads/undergrads/postdoctoral fellows
    I recruit students at all levels in Bayesian statistics, network science and statistical finance. Please contact me by email for availability.

    Research Areas

    • Bayesian statistics
    • Statistics in finance
    • Complex systems science

    Education & Employment Background

    Alex Shestopaloff completed his HBSc (2008) and PhD (2016) degrees in statistics at the University of Toronto, the latter with a thesis on Markov Chain Monte Carlo (MCMC) methods supervised by Radford M. Neal. He then worked in industry at the Canada Pension Plan Investment Board and Vox Pop Labs (2016-2017), returning to academics in 2017 as a Turing Research Fellow in the U.K., mentored by Arnaud Doucet. From 2020 to 2025 he held academic positions in London, U.K. and St. John's, Newfoundland and Labrador, before returning to Ontario to join the University of Guelph.

    Research Themes

    General Description

    My research focuses on 1) the development of novel MCMC methods and Bayesian modelling, 2) the statistical analysis of network data, including detecting when two networks are different, and 3) the development of novel statistical methods for financial applications.

    Current Research Themes

    1. Bayesian online learning. Bayesian online learning is the use of Bayesian inference to incrementally fit a model to a stream of observations. This can be done with the purpose of e.g., performing a forecast. However, real-world processes can generate data streams whose properties change due to changes in the underlying data generation process. Here, the question is: how can we develop methods for fitting models to streams of data that can learn to be adaptable to data that can change over time?

    2. Network science. Networks, on the most broad level, represent data that consists of pairwise relations between some collection of entities (such as a social network, or a network of transactions). It is often of interest to understand whether a network has structure such as communities, or whether it has undergone a change. A question I am interested in is how can we statistically compare two networks to decide if they are different? Extending this, how can we detect a changepoint in a network time series?

    3. Statistical finance. This part of my research studies empirically and develops methods for the analysis of data coming from the financial markets. Some questions of interest here are to do with both low and high frequency data. How can we detect structure, such as lead-lag relationships, in high dimensional ensembles of financial time series data? How can we use high frequency financial data to understand price formation?

    Highlights

    • New Frontiers in Research Fund Exploration Grant 2025-2027, "Improving wellbeing programs’ efficacy: Using adaptive AI to assess and foster wellbeing literacy". $249,978.00. (Co-applicant, PI is Simon Coulombe from Université Laval).