2025 CMS Winter Meeting
Toronto, Dec 5 - 8, 2025
NSERC-CSE Research Communities: Robust, Secure and Safe Artificial Intelligence and Exploratory Analysis of Unstructured Data
Org: Camille Archambault (McGill University), Steven Ding (McGill School of Information Studies) and David Thomson (Tutte Institute for Mathematics and Computing)
[PDF]
Org: Camille Archambault (McGill University), Steven Ding (McGill School of Information Studies) and David Thomson (Tutte Institute for Mathematics and Computing)
[PDF]
- CAMILLE ARCHAMBAULT, McGill University
- BENJAMIN COOKSON, University of Toronto
- SANJEENA DANG, Carleton University
Clustering compositional data with a logistic normal multinomial mixture model with an underlying latent factor structure [PDF]
- BENJAMIN COOKSON, University of Toronto
-
The human microbiome plays a crucial role in health and disease. Advances in next-generation sequencing technologies have made it possible to quantify microbiome composition with high resolution. Clustering microbiome data can uncover meaningful patterns across samples, offering insights into biological variability and disease mechanisms. However, this task presents several challenges. Microbiome data are typically high-dimensional, over-dispersed, and compositional, reflecting relative abundances. As such, analyzing such compositional data presents many challenges because they are restricted to a simplex, which complicates standard statistical analysis. Here, we develop a family of logistic normal multinomial factor analyzers (LNM-FA) by incorporating a factor analyzer structure. The family of models is suitable for high-dimensional microbiome data, as the number of parameters in LNM-FA can be greatly reduced by assuming that the underlying latent factors are small. Parameter estimation is done using a computationally efficient variant of the alternating expectation conditional maximization algorithm that utilizes variational Gaussian approximation. The proposed method is illustrated using simulated and real datasets.
- STEVEN DING, McGill University
- BENOIT HAMELIN, Tutte Institute for Mathematics and Computing
- JOHN HEALY, Tutte Institute for Mathematics and Computing
- TORYN QWYLLYN KLASSEN, University of Toronto
Remembering to Be Fair: Non-Markovian Fairness in Sequential Decision Making [PDF]
- BENOIT HAMELIN, Tutte Institute for Mathematics and Computing
-
Fair decision making has largely been studied with respect to a single decision. Here we investigate the notion of fairness in the context of sequential decision making where multiple stakeholders can be affected by the outcomes of decisions. We observe that fairness often depends on the history of the sequential decision-making process, and in this sense that it is inherently non-Markovian. We further observe that fairness often needs to be assessed at time points within the process, not just at the end of the process. To advance our understanding of this class of fairness problems, we explore the notion of non-Markovian fairness in the context of sequential decision making. We identify properties of non-Markovian fairness, including notions of long-term, anytime, periodic, and bounded fairness. We explore the interplay between non-Markovian fairness and memory and how memory can support construction of fair policies. Finally, we introduce the FairQCM algorithm, which can automatically augment its training data to improve sample efficiency in the synthesis of fair policies via reinforcement learning.
This is joint work with Parand A. Alamdari, Elliot Creager, and Sheila A. Mcilraith.
- PAUL MCNICHOLAS, McMaster University
- GERALD PENN, University of Toronto
- OPENING REMARKS
- KALEB RUSCITTI, University of Waterloo
- GERALD PENN, University of Toronto