Setting risk margin for claims and premium liabilities in accordance with IFRS 17

This proposal deals with the pricing and risk management considerations of a property and casualty (P&C) insurance company. These considerations are within the context of a new accounting standard called IFRS 17, in which liabilities in insurance contracts will be measured prior to and during the exposure periods. We propose an implementable and accurate methodology, which is also compliant with the new standard in generating risk measures and margin adjustments.

Accessible data platform for dynamic experience study of lifestyle underwriting

We seek to replace or enhance the traditional underwriting approach (namely identification of insureds via a pre-defined fixed set of risk criteria) with one based on a set of dynamic protocols that are responsive to human behavioral factors for continual health improvement. We seek to provide a live and interactive in-market research dataset that can be used to explore the benefit of and improve data-driven approaches (namely artificial intelligence or AI) for immediate use in life & health insurance product development and actuarial risk assessment.

Generalization in Deep Learning

In recent years, deep learning has led to unprecedented advances in a wide range of applications including natural language processing, reinforcement learning, and speech recognition. Despite the abundance of empirical evidence highlighting the success of neural networks, the theoretical properties of deep learning remain poorly understood and have been a subject of active investigation. One foundational aspect of deep learning that has garnered great intrigue in recent years is the generalization behavior of neural networks, that is, the ability of a neural network to perform on unseen data.

Advanced Data Science Research for Social Good II

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. Simultaneously, high-quality personnel (HQP) in the domains of data science and urban analytics lack opportunities to work closely with local government to address this gap.

Model development for membership analytics

This project will lay a foundation for Interior Savings Credit Union to better serve their membership by providing them with the means to discover significant membership groups within their large and complex database. This will be done through a combination of developing novel statistical techniques to discover the groups and writing the computer code needed to enact those techniques. We expect the project to benefit Interior Savings and its members by enhancing the credit union’s ability to target growth opportunities.

Deloitte IP Factory: dTRAX Contract Management

Manual contract analysis and management is a laborious task. dTrax is Deloitte’s managed solution to this problem—it uses machine learning to automate the arduous contract management process and help users gain further insight from contracts. Specifically, dTrax will be used to standardize the intake of legal contracts, generate and edit contracts within a web interface, and identify and monitor changes in contracts. The proposed research project aims to improve dTrax by automating contract checking and analysis.

Customer Lifetime Value Prediction Engine: Neighborhood Link Inference and Conversion Prediction

Canada’s financial services industry faces significant challenges to remain internationally competitive in the rapidly evolving web and big data environments. Scotiabank and its global competitors have as a key priority effective use of a large and growing amount of data to optimize the design and pricing of product offerings, to communicate effectively with clients, and to mitigate risk.

Analysis of factors impacting views and revenue of digital assets on YouTube

This research project will focus on factors increasing viewership and revenue on Boat Rocker Media’s digital assets on YouTube. Combination of datasets from various sources such as YouTube revenue reporting, YouTube data and analytics Application Program Interfaces (API’s), and social media will be used in an explanatory analysis and statistical modeling to get new insights. The project will involve integration of data from various sources as well as data cleansing.

Prediction of Optimal Business Structure for Tax Efficiency

In the Corporate Tax domain, professionals must review hundreds of documents in the process of filing taxes and rely on experience to identify where benefits or deductibles can be applied. This manual task is subject to human error and can result in unnecessary administrative overhead. With access to PricewaterhouseCoopers’ wealth of tax data, it is possible to develop a tool that uses historical trends to automate this process.

Open-domain Contextual Conversation Generation

The objective of the project is to design a system that is able to generate context-wise reasonable and meaningful responses to open-domain conversation queries. In open-domain conversation generation, the retrieval-based methods and neural network generative models are two main approaches; there are also some recent research about improving the context consistency of conversation generation.