Complex Network Data Analysis for Systemic Risk Management and Loss Prevention

Financial fraud is a very serious problem plaguing financial institutions. Recent advances in information technology have only exacerbated this problem. Poor risk models were at the core of the 2008 financial crisis. Complex networks, structured graphs, are powerful models for representing higher-order interactions or dependencies within data sets. Graphs are relational models of covariates, where nodes represent variables and arcs their connections (relationships). Based on our previous theoretical work in the field, we aim to apply complex network (graph) techniques to the area of fraud prevention and risk management. The end goal of this project will be to develop a portfolio of analytical tools to reduce financial losses due to fraud, credit defaults and market fluctuations.

Faculty Supervisor:

Yuri Lawryshyn;Cristian Bravo Roman

Student:

Partner:

Royal Bank of Canada

Discipline:

Mathematics

Sector:

Finance and Insurance

University:

University of Toronto

Program: