With the number of credit card transactions rapidly increasingly by the day, University of Ottawa students are doing what they can to tackle credit card fraud with advanced machine learning technology.
Bahar Emami Afshar, a master's student in computer science, has created an advanced machine learning program designed for the financial sector that can detect anomalies in large datasets.
“Credit card fraud has become a key issue in today's world due to the rapid growth of e-commerce and online trading,” Afshar said in a media release. “What's even more urgent is that this type of fraud is often linked to a wider range of illegal activities, such as identity theft and money laundering.”
According to Canada's payments, in 2022, Canada saw a 20.5 billion payment transaction. This scale of dataset makes it nearly impossible to manually identify.
AFSHAR's Iterade (Iterative Anomaly Detection Ensemble) tool prioritizes labeling efforts to maximize fraud detection.
Testing of the Iterade tool has shown preliminary success. Afshar and her collaborators have seen fraud detection rates of 3-15 times better as a result of this tool, as a result of previous tactics.
Credit card fraud remains a critical issue facing the business sector, so its success is important. The Cainments Canada report released in September 2024 has been published over the past six months. One in five Canadian companies I had experienced payment scams. 20% of these cases were credit card fraud, the third major cause.
“Dealing with the risk of fraud is a central focus of the payment ecosystem. It requires a multifaceted approach that leverages technology, system innovation, evolving regulations and education through ongoing industry cooperation,” Canada's Head of Payments Donna Kinoshita said in a press release.
That's exactly what AFShar tools strive for, and its success could prove beneficial for other sectors as well.
“In the research community, Iterade introduces a new, unsupervised anomaly detection approach that shifts how to deal with highly imbalanced datasets. It provides a flexible, budget-independent framework that outweighs existing methods of fraud detection, and sets new standards for handling other imbalanced classification problems,” Afshar said.
As her work continues, AFSHAR and her team aim to integrate Iterade into a proactive learning framework that can solve challenges across the financial sector in industries such as healthcare, cybersecurity and infrastructure monitoring.
“We look forward to pursuing a career combining AI and engineering to tackle complex real-world problems,” Afshar says. “I aim to contribute to cutting-edge AI applications that improve decision-making, security, and overall quality of life across a variety of industries.”
