Isaac Rankin, a PhD student in statistics at the University of British Columbia, is focusing his research on exact inference with approximate computational methods. His work sits in the Faculty of Science and is tied to a broader push to make advanced statistical methods both theoretically sound and practical for scientific use.
Key facts
- Rankin is pursuing a Doctor of Philosophy in Statistics at UBC.
- His research topic is exact inference with approximate computational methods.
- He is supervised by Saifuddin Syed and Charles Margossian.
- He has been selected for the Canada Graduate Research Scholarships Doctoral (CGRS D) Program and the Four Year Doctoral Fellowship (4YF).
- He is from the Comox Valley in Canada.
What this means for graduate students and applicants
Rankin’s story highlights the kind of research environment UBC presents to doctoral students in statistics: one that combines Bayesian and computational methods with research that can be implemented and used more broadly. For applicants interested in this area, the program example shows that UBC supports work on advanced methods such as enhanced sampling, variational inference and generative models.
The practical takeaway is that students who want to work at the intersection of theory, computation and application may find a fit in UBC’s statistics PhD. Rankin says the department’s faculty, along with tools such as Stan, GPyTorch and Pigeons.jl, helped make the program attractive because of its mix of theoretical work, implementation and collaboration.
- Program: Doctor of Philosophy in Statistics (PhD)
- Faculty: Faculty of Science
- Supervisors: Saifuddin Syed and Charles Margossian
- Selected awards: CGRS D and 4YF
Why did Rankin choose UBC?
Rankin says he chose UBC because he considers it the strongest statistics department in Canada for Bayesian and computational methods. He also points to Vancouver’s access to cycling and skiing as part of the appeal, describing it as a good personal fit.
Before coming to UBC, he studied at Oxford and worked on diffusion models for molecular generation during his master’s degree. He says undergraduate NSERC research terms and his Oxford degree prepared him for productive research at UBC.
What is his research about?
Rankin’s research addresses a common problem in scientific computing: exact methods can be too slow, while approximate methods can be faster but may lose guarantees. His work looks at how to recover those guarantees through provably correct approximations and through methods that tolerate misspecification while still preserving the key properties needed for downstream decisions.
He says his main interests are the theoretical guarantees themselves, along with practical methods for enhanced sampling, variational inference and generative models.
What does the program offer students?
According to Rankin, UBC’s statistics department has several faculty members working on Bayesian and computational methods. He also highlights the department’s emphasis on making its methods widely available through software and collaboration with other research groups.
For students, that means the program is not only about theory. It also appears to offer exposure to implementations that can influence scientific work beyond the department itself.
Frequently asked questions
What is Isaac Rankin studying at UBC?
He is studying for a PhD in Statistics at UBC, with research focused on exact inference using approximate computational methods. His work aims to find ways to preserve useful guarantees while still taking advantage of faster approximate techniques.
Which awards has Isaac Rankin received?
He has been selected for the Canada Graduate Research Scholarships Doctoral (CGRS D) Program and the Four Year Doctoral Fellowship (4YF).
