We’re here to clear the mist around AI

We Are Recruiting!

Multiple positions available starting fall 2019.

Join Us


Talk Series01/28/2020 17:00:00

Demystifying How Machines Learn

The series will introduce the methodology of human psychology insights to solve unsupervised learning problems and build novel machine learning and mathematical models. The series is open to all audiences, especially those who are passionate about unsupervised learning and reinforcement learning. Our speaker, Nishkrit Desai, will lead this series. He is a student researcher involved in projects with organizations such as Google, Tesla, Autodesk, and OpenAI. His main interests are unsupervised learning and reinforcement learning. Join us in uncovering exciting ideas in machine learning!

Talk Series11/12/2019 18:00:00

Research Talk - Professor Pascal Poupart

“Unsupervised Video Object Segmentation for Deep Reinforcement Learning”, by Professor Pascal Poupart. He will discuss a new technique for deep reinforcement learning that exploits flow film information, detects moving objects and uses their information for action selection.

Forum9/22/2019 8:30:00

The Second AI-Squared Forum

AI Squared is a forum for facilitating dialogues across boundaries in the AI industry. It is a great opportunity to - Discuss big ideas with AI researchers working on cutting-edge projects - Network with recruiters from AI companies such as Uber ATG and Shopify to secure internship opportunities - Hear about the experience of undergraduates working at institutes such as the Vector Institute, Nvidia and Google - Connect with AI professors and professionals from academia and industry

Meeting9/16/2019 18:00:00

Annual General Meeting (AGM) 2019

Come and meet the new executive team! We will be talking about a variety of events that will be held throughout the year. At the end of the AGM, project leads will also introduce new academic projects. They will meet with anyone interested in joining the project teams afterwards. Anyone applied/interested in applying for a project associate position is strongly encouraged to come and potentially find your team.

Paper Reading4/25/2019 18:30:00

Exploration in Reinforcement Learning

Sicong (Sheldon) Huang is currently an undergrad at U of T. His research interests include machine learning and cognitive science. Currently he is a research intern at Vector Institute and Borealis AI, where he works on evaluating generative models. Previously he worked on deep generative models for domain transfer on text and music, such as TimbreTron (ICLR2019) and CipherGAN (ICLR2018). Currently he is interested in some fundamental questions regarding learning and cognition. He also co-founded two organizations in his free time including UTMIST and FOR.ai, an international distributed AI research collaboration. He will be giving a broad introduction on exploration strategies in Reinforcement Learning. Exploration vs. exploitation trade-off is one of the core problem in modern deep reinforcement learning especially when the environment is complex and partially observable. Exploiting the best actions based on the current policy can yield a very good short-term return, but can easily get the agent stuck in a local minimum. The goal of this paper reading group is to demystify the principle behind the deep RL exploration problem. The topics of this paper reading group range from optimistic exploration, posterior sampling to information gain in exploration. In addition to those traditional approaches, Sheldon will also be going over several state-of-the-art approaches in the field.

Talk Series3/12/2019 18:30:00

Using ML Method on Assessing Drivers' Vigilance

Speaker: Min Liang Min is currently working as a data scientist at RBC after receiving her master’s degree in electrical engineering from McGill University. She has had fruitful research experience in ECE from Tianjin University, Harvard University, MITACS and McGill University. This talk session will be featuring her MITACS research project at Alcohol Countermeasure System Corp., an alcohol tester design company, in 2017. The overall objective is to reliably assess drivers’ vigilance using non-intrusive measures. The data used for this project are collected by Smart Eye Pro, an eye-tracking system. The research project mainly focuses on identifying the most important features and building classification/regression systems to reliably detect drowsiness and fatigue. Join us on March 12th. Get inspired, make connections and learn new skills from this talk session!

Talk Series3/5/2019 18:30:00

The Effect of ASR on Alzheimer's Detection

Ever wondered how Machine Learning works for medical diagnostics? If so, don’t miss this insightful guest speaker talk given by Jekaterina Novikova. We are honoured to invite her to talk about the effect of ASR (automatic speech recognition) on classification performance in Alzheimer's detection. Jekaterina is a director of Machine Learning at Winterlight Labs, where they focus on developing a novel AI technology that can quickly and accurately quantify speech and language patterns to help detect and monitor cognitive and mental diseases. Previously, she received her PhD from University of Bath. She also worked as a post-doctoral researcher at Heriot-Watt University in Edinburgh, UK, at the Interaction Lab and Natural Language Processing Lab. Join us on March 5th for this insightful speaker session and learn from this excellent machine learning researcher!

MIST1012/5/2019 18:30:00

Computer Vision

This workshop covers basic knowledge of recent advances in computer vision. We’ll introduce the commonly used tools and models in solving computer vision tasks and explore how the frontier computer vision research is accelerating integration to real-life scenarios. Following that, we will demo a typical workflow for CV projects.

Talk Series11/27/2018 18:30:00

Machine Learning for Recommender Systems

Recommender systems are intelligent machine learning systems that help customers discover personalized products from a dynamic pool of diverse choices. In this talk, Soon Chee Loong from Data-Driven Decision Making Laboratory will share his insights on different types of recommender systems, their challenges, and various seminal approaches to tackle them.

Talk Series11/20/2018 18:30:00

Detecting Alzheimer's Disease

We had the honour to invite Jekaterina Novikova to talk about Machine Learning Methods in Detecting Alzheimer's Disease from Speech and Language. Jekaterina is a Director of Machine Learning at Winterlight Labs, where they focus on developing a novel AI technology that can quickly and accurately quantify speech and language patterns to help detect and monitor cognitive and mental diseases.

MIST10111/19/2018 18:30:00

Reinforcement Learning

Lunjun Zhang will be providing an enriched introduction to reinforcement learning, one of the most active fields of AI research today. The workshop will cover both the fundamental ideas and several state-of-the-art methods including value iteration, policy gradients, actor critic, Q learning, and exploration strategies. The workshop will also explore how concepts from optimisation and information theory are used for control.

MIST10111/8/2018 18:30:00

Introduction to Machine Learning

A general introduction to machine learning fundamentals and a demonstration of typical workflow in solving machine learning problems. The typical workflow is demonstrated here: https://youtu.be/4flbCsGBicE

Talk Series10/30/2018 18:30:00

Machine Translations

We had the honour to invite Jun Gao to talk about Recent Advances in Neural Machine Translation. Jun Gao graduated from Peking University with fruitful experiences gained from working in both academia and industry. Currently, he is a graduate student in the Machine Learning Group here at UofT.

Paper Reading10/23/2018 18:30:00

Using Deep Learning to Analyze Typhoon Images

Speaker: Danlan Chen Machine Learning Researcher at Borealis AI, the machine learning research department at RBC Master’s from McGill University, where she did research in Machine Learning in the Reasoning & Learning Lab under the supervision of Professor Doino Precup. Check out her personal website: https://danlanchen.github.io/

Paper Reading10/9/2018 18:30:00

Deep Learning on Graphs: CNN, RNN, GNN

We are hosting the first session of our research paper reading group series on Graph Neural Net on Computer Vision Problems with Arie Huan Ling from NVIDIA AI Research lab and Vector Institute as our guest speaker!

Meeting9/13/2018 18:30:00

Annual General Meeting (AGM) 2018

UTMIST’s mission is to clear the mist around Machine Learning and create a platform of opportunities for UofT students to get involved in the ML community in Toronto. This year we are hoping to bond with more AI/ML enthusiasts, no matter with previous experience or not. MIST101 will stay with a re-structured format for new-comers while more hands-on projects and deeper study with researchers on state-to-art techniques are provided.

2018 Events

About Us

University of Toronto Machine Intelligence Student Team (UTMIST) is an officially certified student organization within the University of Toronto. Our mission is to let more people get to know about artificial intelligence and “clear the mist” around it!




Find Us On