Advanced analytics and predictive statistics in continuous flow sports.

Hockey has long been shown to be among the least predictable of all professional sports. Recent developments in data collection methods have created the demand for more detailed and advanced predictive modelling techniques to extract value from and apply the data to real world problems. This project focuses on predicting important outcomes in hockey at both team and player levels. Game winners and scores will be predicted using Bayesian approaches tailored to accommodate evaluative statistics and relevant pre-game factors.

Learning Jungle AI Recommender System for Enhanced Education

Devhaus Corporation operates 20 early child education centers across 5 countries including Canada, USA, Singapore, Cambodia and Philippines. To further implement its principle of Observation to Education, Devhaus is partnering with York University research team to develop an Artificial Intelligence Recommender System to help teachers across the world select optimal lesson plans for each kid based on the learner's behaviors.

Optimization of Savings and Retirement for Canadians

There are numerous financial goals that most Canadians face. Retirement, funding post high school education, managing debt, purchasing appropriate amounts of insurance and saving for lump sum purchases. Each of these goals has various accounts and savings vehicles associated with them. The research projects we are proposing will help Canadians define their own financial situation, focus on their goals in the optimal order, and best utilize savings vehicles and government benefits to best meet their goals. Glencairn Financial Inc.

Improving energy system planning solutions by accounting for inherent uncertainties through robust optimization

More and more distributed energy resources (smart loads, self-generation, electric vehicles, etc.) are installed directly at the customers.  This causes fluctuations in the distribution network that can reverse the power flow or increase the cold pick-up effect.

An Artificial Intelligence algorithm for creating personalized learning journeys for students

With the fast growing information available on the web, students are often greeted with countless learning materials. As such, personalization is an essential strategy for facilitating relevant learning materials to satisfy students’ needs. The scope of this project is to design a recommendation system by using a deep learning process for personalized learning based on a quiz module. At the end of the project, we would be able to determine how students like to learn and to evolve the learning path based on strengths to enhance the learning experiences.

Optimization of group equivariant convolutional networks

The explosion of popularity of deep learning owes a lot to the success of convolutional neural networks, widely used in diverse fields including computer vision and natural language processing. Recently, the group equivariant convolutional neural network (G-CNN) was introduced, where equivariance of symmetries inherent in the data set is built in the architecture of the networks.

Augmented reality immersive simulation for flight deck design and evaluation “Holodeck”

The aircraft flight deck has increased substantively in its complexity in recent years. The input systems are more complex, and the information feeds are much more detailed. In order for a pilot to interface effectively with the aircraft systems, the cockpit control functions must be laid out in an intuitive format. To do this, a trial and error approach is required, with meaningful input at each design phase.

Leverage on Artificial Intelligence in Capacity Management: Predict IT assets usage based on Business events

The Societe Generale Bank possesses a network of trading applications which generate the hardware consumption data (CPU, memory and network communication).

Signal recognition with machine learning using wavelet features

The emerging techniques of machine learning and artificial intelligence are making revolutionary changes in all kinds of the industrial world. As a high-tech business solution company, uses these modern techniques to help industrial manufactory companies work more efficiently. One of the challenging problems is to make the computer automatically recognize the status and behavior of the machine from the data collected by different sensors, so that people can record the history of the machine and conduct further analysis.

Mobile Mesh Technology For Improved Connectivity In Canada

Lack of affordable, reliable access to the Internet in remote, rural, and Indigenous communities in Canada and internationally has led to a digital divide affecting billions. Left has developed its RightMesh technology, which lets nearby mobile devices connect without the Internet. Normally data flows from a phone, through Internet Service Providers (Telus, Shaw), to the cloud (Google, Amazon), and back to reach another person, despite being nearby. With the RightMesh network, data can flow from phone to phone to phone until it reaches the intended person.