Mark Your Calendar for the CANSSI/SSC Cross-Country Tour

> français

Want to know what your colleagues in Canada’s statistical science community are working on these days? As part of the 50th anniversary celebrations of the Statistical Society of Canada (SSC), CANSSI and the SSC have partnered to organize a CANSSI/SSC Cross-Country Tour with short seminars hosted at universities from coast to coast.

Each session will feature a guest speaker talking about their research on Zoom (and in some cases also in person). Sessions will be hosted by Joanna Mills Flemming, Associate Director, CANSSI Atlantic, and Professor in the Department of Mathematics and Statistics at Dalhousie University.

We hope you’ll be able to grab a cup of coffee and join us!

Up Next

The Development of a COVID-19 Self-Assessment Risk Model to Mitigate Case Importation Risk in Yukon

Dr. Lisa Kanary, Instructor, Mathematics & Statistics, School of Business and Leadership, Yukon University
Hosted by Dr. Joanna Mills Flemming, CANSSI Associate Director, Atlantic Region

Friday, September 30 | 12:00–12:45 p.m. MST
Yukon University, Whitehorse, Yukon
On Zoom

REGISTER HERE

Presentation abstract: Most travel restrictions imposed during the COVID-19 pandemic applied to all travellers, regardless of origin of departure or individual behaviour. Such measures do not account for the role that behaviour plays in an individual’s probability of encountering an infectious individual and contracting COVID-19. We present a statistical model, designed for use in the Yukon Territory, that estimates an individual’s probability of being a contact of an infectious person as a function of disease prevalence and the daily activities the individual engaged in during the two weeks prior to the date of assessment. Our tool presents a method for estimating contact probability that could be adopted by jurisdictions considering border travel restrictions, facility closures or group size limits, or for individuals evaluating their own behaviours.

About Lisa Kanary: Dr. Kanary is an instructor in the School of Business and Leadership at Yukon University in Whitehorse, Yukon. She also coordinates the Bachelor of Business Administration degree program that Yukon University offers.

Originally from Nova Scotia, Dr. Kanary moved to the Yukon in 2013. Since arriving in the north, she has spent time in many different areas of the University, such as instructing in the Schools of Science, and Academic Skills and Development, developing the Climate Change certificate, performing research at the YukonU Research Centre. Wearing many hats at the University has given her a unique perspective of the many avenues and opportunities students can pursue.

Currently, Dr. Kanary’s background is in applied math. She is the primary business statistics knowledge holder and teaches statistics classes in the Bachelor of Business Administration. Lisa infuses her classes with real world experience by partnering students with community groups and industry to work through statistical problems and challenges. Lisa also continues to be active in research as a data scientist and math modeller. She is currently working on math models that pertain to COVID-19 with a number of colleagues locally and around the globe. She is excited to support YukonU research and students on their journey forward.

Past Sessions

Sparse Hamiltonian Flows (or: Bayesian Coresets Without All the Fuss)

Dr. Trevor Campbell, Assistant Professor, Department of Statistics, University of British Columbia
Hosted by Dr. Joanna Mills Flemming, CANSSI Associate Director, Atlantic Region

Friday, July 8 | 12:00–12:45 p.m. PDT
University of British Columbia
On Zoom

WATCH THE WEBINAR RECORDING

Presentation abstract: Bayesian inference provides a coherent approach to learning from data and uncertainty assessment in complex, expressive statistical models. However, algorithms for performing inference have not yet caught up to the deluge of data in modern applications. One approach—Bayesian coresets—involves replacing the large dataset with a small, weighted, representative subset of data during inference. The coreset is designed to capture the information from the full dataset, but be much less computationally expensive to store in memory and iterate over. Although the methodology is sound in principle, efficiently constructing such a coreset in practice remains a significant challenge: current methods tend to be complicated to implement, slow, require a secondary inference step after coreset construction, and do not enable model selection. In this talk, I will introduce a new method—sparse Hamiltonian flows—that addresses all of these challenges. The method involves first subsampling the data uniformly, and then optimizing a Hamiltonian flow parametrized by coreset weights and including periodic momentum quasi-refreshment steps. I will present theoretical results demonstrating that the method enables an exponential compression of the dataset in representative models, and that the quasi-refreshment steps reduce the KL divergence to the target. Real and synthetic experiments demonstrate that sparse Hamiltonian flows provide accurate posterior approximations with significantly reduced runtime compared with competing dynamical-system-based inference methods.

This talk will be based on two papers that are available online as preprints:

You Are What You Eat: Advances in Marine Predator Diet Estimation via Fatty Acids

Dr. Connie Stewart, Professor, Department of Mathematics and Statistics, UNB
Hosted by Dr. Joanna Mills Flemming, CANSSI Associate Director, Atlantic Region

Friday, June 17 | 12:00–12:45 p.m. ADT
University of New Brunswick at Saint John
On Zoom

WATCH THE WEBINAR RECORDING

Presentation abstract: In marine ecosystems, estimating predator diets can be especially challenging since feeding cannot typically be directly observed. In this context, quantitative fatty acid signature analysis (QFASA) was devised as an indirect method of estimating predator diets and has successfully been applied to a variety of seabird species, marine mammals and fish. The approach estimates the proportion of each prey species in a predator’s diet by matching the predator and prey fatty acid signatures.   

From a statistical methodology and analysis perspective, diet estimation by way of fatty acid data present several challenges, primarily because both the fatty acid signatures and diet estimates represent compositional data, but also due to a variety of practical considerations. In collaboration with biologists at various institutions across Canada, my research has focused on extensions to QFASA, inference procedures for predator diets, and the development of new probabilistic models for diet estimation through fatty signature analysis. In this talk, I will highlight some recent advances and ongoing work in this area.

Cars, Steaks, and Hurricanes: A Bayesian Approach to Inverse Problems for Random Vectors

Dr. Don Estep, Scientific Director, CANSSI
Hosted by Dr. Joanna Mills Flemming, CANSSI Associate Director, Atlantic Region

WATCH THE WEBINAR RECORDING

Presentation abstract: Scientific inference and engineering design frequently involves the inverse problem of determining information about the state of a complex physical system from observed data of its behaviour. I first encountered this kind of inverse problem in graduate school with somewhat disastrous results. The “after action” analysis of that experience had several consequences for my career, including a sustained effort in the formulation and solution of inverse problems for physics and engineering models. I will give an intuitive description of our Bayesian formulation and solution of inverse problems in the context of cooking steaks and forecasting of hurricane storm surge. I will also talk about extensions of our approach and relation to other Bayesian statistics.

Future Dates

Friday, September 30 | 12:00–12:45 p.m. MST | Yukon University, Whitehorse, Yukon
Title to be announced
Dr. Lisa Kanary, Instructor, School of Business and Leadership
Applied Science and Management

Friday, October 14 | 12:00–12:45 p.m. EDT | McMaster University, Hamilton, Ontario
Title to be announced
Dr. Lehana Thabane, Professor and Interim Chair, Department of Health Research Methods, Evidence and Impact

Friday, November 18 | 12:00–12:45 p.m. EST | Biogen, Montreal, Quebec
Title to be announced
Dr. Gabrielle Simoneau, Senior Principal Biostatistician

Posted in News | Comments Off on Mark Your Calendar for the CANSSI/SSC Cross-Country Tour

Comments are closed.