Welcome to our casual research seminar organized in the Department of Statistical Sciences at the University of Toronto. Our aim is to discuss new ideas with everyone in the department (students, postdocs, visitors, and faculty). Talks usually last around 30 minutes, followed by discussions. We welcome current research, overviews of emerging topics, statistics pedagogy, research methodologies, and more. Some pizzas will be offered before the seminar around 12:20pm.
Seminar organizers: Christopher Blier-Wong & Biprateep Dey & Dehan Kong & Kathleen Miao & Chao Wang
Meetings are on Tuesdays, 12:30pm (30 min talk + discussion), room: 9014, Ontario Power Building (700 University Avenue). Pizza will be served around 12:20pm.
To sign up to give a talk, use our spreadsheet.
Upcoming talks
| Date | Speaker | Title | Remarks |
|---|---|---|---|
| Sep 16 | Rohan Alexander | Two Applications of LLMs in Data Science | Overview of two works in progress: 1) comparing different LLMs on a specific data science analysis, 2) comparing LLMs on a broad data science analysis |
| Sep 23 | Skip for “The Fast and Curious 2: MCMC in Action” Conference | ||
| Sep 30 | Skip for CANSSI Ontario Distinguished Lecture in Statistical Sciences by Nancy Reid | ||
| Oct 07 | Liam Welsh | Intro to Optimal Execution | A broad and approachable introduction to limit-order books and trade execution in financial markets, based off this paper |
| Oct 14 | Sam Caetano | Statistics Outreach | Different opportunities to connect with different audiences and how to approach data literacy and civic statistics with young students and/or general audiences |
| Oct 21 | Sabrina Sixta | How to Solve It | A summary of Polya’s book and a broader discussion on how to solve problems and generate novel results |
| Oct 28 | READING WEEK BREAK | ||
| Nov 04 | Thibault Randrianarisoa | Gaussian Variational Predictive Posteriors of Wide NNs Often Resemble the Prior | In wide overparameterized neural networks, using a Gaussian variational posterior to approximate the parameter posterior hinders the predictive posterior’s ability to learn from data |
| Nov 11 | Ricardo Baptista | Data Memorization and the Need for Regularization with Score-Based Diffusion Models | |
| Nov 18 | Konstantinos Christopher Tsiolis | High-Dimensional Learning Dynamics: Recent Progress and Open Problems | Present recent theoretical studies of stochastic gradient descent (SGD) training of linear models and neural networks on high-dimensional problems |
| Nov 25 | Chris J. Maddison | Scaling Laws - An Opportunity for Statisticians | Present some of the empirical developments in the scaling behaviour of large language model training and propose that this is a rich set of problems for statisticians to consider |
| Dec 02 | Mehdi Molkaraie | Graphical Models and Transformations | |
| Dec 09 | Jianhui Gao | Reliable Inference for Biomedical Research | |
| Dec 16 | HOLIDAY BREAK | ||
| Dec 23 | HOLIDAY BREAK | ||
| Dec 30 | HOLIDAY BREAK | ||
| Jan 06 | Marco Antonio Gallegos Herrada | Bayesian Inference for HMMs | Present the work developed for Bayesian inference for HMMs under multimodality with an application to ecological time series |
| Jan 13 | Morris Greenberg | The Past, Present, and Future of Scrabble Engines | I’ll be taking a step back from my usual research to talk about something a bit different: how computers learn to play Scrabble. |
| Jan 20 | Room unavailable (MScAC internship expo) | ||
| Jan 27 | TBD |
Past talks
| Date | Speaker | Title | Remarks |
|---|---|---|---|
| May 6 | Yuan Tian | Leveraging multimodal neuroimaging and GWAS for identifying modality-level causal pathways to Alzheimer’s disease | Internal Speaker |
| Apr 29 | Andrey Feuerverger | Statistical Theory and the Primes (49min talk) | Professor Emeritus at DoSS |
| Apr 22 | Mauro Filomeno | Online Preference Optimization for Diffusion Models with Classifier-Free Exploration | Student supervised by Qiang Sun |
| Apr 15 | Sohee Goo & Ricky Chen | AI-assisted grading | Students supervised by David Liu (CS) and Nathan Taback |
| Apr 8 | Aram-Alexandre Pooladian | Wasserstein Flow Matching: Generative modeling over families of distributions | External Speaker |
| Apr 1 | Liam Welsh | Multi-Agent Reinforcement Learning for Greenhouse Gas Offset Credit Markets | Internal Speaker |
| Mar 25 | Jianhui Gao | Using ML predictions to drive large-scale and robust scientific inquiry | Internal Speaker |
| Mar 18 | Ziyi Liu | Sequential Probability Assignment with Contexts: Minimax Regret, Contextual Shtarkov Sums, and Contextual Normalized Maximum Likelihood | Internal Speaker |
| Mar 11 | Ichiro Hashimoto | Universality of Benign Overfitting in Binary Linear Classification | Internal Speaker |
| Mar 4 | Skip for UTSC search job talk | ||
| Feb 25 | Zhijing Jin | Causal Reasoning with Large Language Models | Incoming Assistant Professor at CS |
| Feb 18 | Skip for reading week | ||
| Feb 11 | Skip for UTSC search job talk | ||
| Feb 4 | Skip for job talk | ||
| Jan 28 | Skip for job talk | ||
| Jan 21 | Skip for job talk | ||
| Jan 14 | Sabrina Sixta | MCMC common random number simulation: discoveries and mysteries | Internal Speaker |
| Jan 7 | Jeffrey Rosenthal | Chess champions, CEOs, streaks, and probabilities | Internal Speaker |
| Dec 10 | Claire Yu | Speed up your computation: strategies and platforms | Internal Speaker |
| Dec 3 | Ruyi Pan | Devariation: a robust approach to improve statistical power in high-dimensional multi-view association testing | Internal Speaker |
| Nov 26 | Kevin McKinnon | BP3M: Bayesian Positions, Parallaxes, and Proper Motions derived from the Hubble Space Telescope and Gaia data | Astro postdoc |
| Nov 19 | Kathleen Miao | Robust Elicitable Functionals | Internal Speaker |
| Nov 12 | Skip for DoSS Postdoc Day on Nov 14 | ||
| Nov 5 | Nancy Reid | Some aspects of inference under model misspecification | Internal Speaker |
| Oct 29 | Jun Young Park | What makes good applied statistics research? Some recent case studies from neuroimaging statistics | Internal Speaker |
| Oct 22 | Piotr Zwiernik | Property testing in Gaussian graphical models | Internal Speaker |
| Oct 15 | no meeting | ||
| Oct 8 | Scott Schwartz | Incorporation of AI chatbots into STA130 | Internal speaker |
| Oct 1 | DSI Research Day | Skip | Skip |
| Sep 24 | Vianey Leos Barajas | Statistics Summer Research Clubs (AI for Baseball, Shark Statistics) | Internal speaker |
| Sep 17 | Biprateep Dey | Calibrated Uncertainty Quantification for Physical Sciences | Internal speaker |
- We would like to acknowledge funding from the Department of Statistical Sciences.