You are here

Statistical Learning of Complex Data with Complex Distributions

Project Type: 
Past

The project addresses some of the main practical and theoretical difficulties encountered when dealing with large datasets.

Project Leader(s): 

Dr. Yoshua Bengio, Université de Montréal

Statistical machine learning is an endeavor in which statisticians and computer scientists use computation to identify useful information from large amounts of data. Telecommunications, insurance and pharmaceutical companies use the team’s machine learning and data mining techniques to determine customer patterns, predict future customer behavior and better understand their needs. The project addresses some of the main practical and theoretical difficulties encountered when dealing with large datasets. In the past year, the team focused on algorithms that learn to represent data at multiple levels of abstraction, and discovered new algorithms for learning each level given the previous one. Reinforcement learning algorithms were discovered by letting a computer play the Chinese game of Go against itself for a very large number of games. This process enabled the computer to learn how to play the game. The team also found ways to simplify the difficult optimization task involved in large classes of learning algorithms.

Project team: 
Dr. Hugh Chipman, Acadia University
Dr. Dale Schuurmans, University of Alberta
Dr. Pascal Vincent, Université de Montréal
Dr. Shai Ben-David, University of Waterloo
Funding period: 
February 25, 2022 - March 31, 2021