Anomaly Detection in Financial Data

In this joint collaboration with Scotiabank we hope to solve a commonly faced problem by large financial institutions. It is to detect errors in financial datasets. This could be due to typing errors made by a human or a computer glitch that causes an incorrect value to be stored. To identify these errors, we plan to build an error detection system. It will model how financial variables change in relation to other variables. This will help us identify groups of variables that move, through time, in a similar manner. With this knowledge we will then be able to spot errors in the data.

Investigating Insurance Insolvency in Canada's Property and Casualty Industry

PACICC role is to compensate policyholders in scenarios where a P&C insurer can no longer provide compensation while overseeing the health of the P&C industry in Canada. The proposed project aims to improve PACICC’s ability to identity companies at risk of insolvency and improve strategies to minimize dead weight loss when insolvency is imminent.

Intraday Trading and Analysis and Monitoring Trader Behavior

Electronic exchanges are venues that provide immediacy for those who need to find a counterparty to their trades. Orders of various types arrive in the market at ever increasing speeds, and in this era of high-frequency trading (HFT), institutional investors are often disadvantaged because of their high-latency relative to faster traders.

Risk aggregation beyond the normal limits

Risk aggregation is omnipresent in insurance applications. A recent example, borrowed from the modern regulatory accords, is the determination of the aggregate economic capital and its consequent allocation to risk drivers. A more traditional illustration of the importance of risk aggregation in insurance is the celebrated collective risk theory that dates back to the early years of the 20th century. This project will assist Sun Life Financial to build and implement an efficient quantitative framework to approximate the aggregate risk of its portfolio.

Risk Margin for Claims and Premium Liability in Accordance with IFRS 17

The Building Block Approach (BBA) is one of the liability measurement approaches proposed in the new insurance contract standards - International Financial Reporting Standards (IFRS) 17. Of the three components under BBA, determining the risk margin is the most essential.

Evaluating offline functionality of progressive web applications in e-commerce business: A/B testing and causal models

The project is to understand the potential impact of enabling the offline functionalities on progressive web application utilized as e-commerce platforms. For instance, will enabling offline functionalities help increase revenue by providing a more engaging environment when the connection is poor? Traditional evaluation method is to run online controlled experiments by randomly assigning users into two groups: with and without the feature. However, to accurate assess the impact, one needs to identify appropriate key metrics to quantify the performance.

Advancing an Artificial Intelligence Platform for Crop-Health Monitoring

Plants can respond to changes in their surroundings and can convey precise information about their health state. Ecoation has developed a multi-sensory data acquisition device to capture this information and has been collecting in-field sensor data along with data labels produced by human experts during data collection. In addition, images of various parts of plant canopy has also been collected to supplement the sensory information and to provide insights into plant physical features such as vegetation.

Wind Turbine Power Curve Modelling for Reliable Power Prediction Using Isotonic Regression and Different Loss Functions

Electrical power generation based on wind energy has been one of the fastest growing renewable energy sources. An important area of research in wind energy is to find different ways to improve the power reliability of systems. Modeling wind turbine power curve using past data is often used as an efficient way to use empirical power curve instead of manufacturer company power curve.
As wind-power data are often so noisy, fitted wind turbine power curves could be very different from the theoretical ones that are provided by manufacturers.

Advancing Data Science Research for Social Good

Due to rapid development of technology, such as the Internet of Things, collecting data is easier and cheaper than ever before. As a result, municipal governments and urban centres across Canada are being inundated with data—data that have potential to improve public service. Despite this, local governments do not have enough data expertise to extract insight from these overwhelming datasets, which are often unstructured and “dirty” (i.e., incomplete, inaccurate, and/or erroneous).

Creating moments for shoppers: Impact of time on effectiveness of notifications

Over the last recent years, internet usage through mobile devices has grown rapidly and today, majority of the online traffic is coming from mobiles. This means that most of the time, a user uses his/her mobile to view the retailer website, review items or finish a purchase. Accord ingly, reta ilers have come to the idea of building a platform that would engage and re-engage users through pushing notification. However, there are multiple factors such as time of a day that a retailer would takes into account to send out notific.ation.