The data analysis company FICO wrapped up its semester-long contest with Morehouse College computer science students to encourage a more diverse future in data science. They taught students how to use AI software to design detection models banks could use to fight fraud.
To promote diversity, equity, and inclusion (DEI) in data science, The FICO Educational Analytics Challenge was created for Historically Black Colleges and Universities (HBCU) students to give real-world scenarios in financial fraud detection featuring anomalies and features in a real-time fraud detection model.
The judges named the team “FinPaid” as the winners of the challenge. Their team captain, Mark Ross, a junior Computer Science major from Atlanta, said the competition involved learning how to build a predictive fraud detection model from start to finish. The executives and analysts taught the students the Python programming language needed to build their models.
A fourth-year student from Chicago and captain of team “Data Wizards,” Julius Jones was inspired by what he learned from competing in the challenge. He said he wanted to learn more about fraud analysis so he could be ready when he graduates.
Scott Zoldi, Chief Analytics Officer for FICO, enjoyed the enthusiasm of the students and their professors in preparing them for a role in the tech industry.
“It’s great to be here and to have that relationship so that we can empower and actually accelerate the graduates of those programs being effective on day one so that they’re well aligned with no impedance match between what industry needs and what they’re learning,” Zoldi said.
He helped create the competition to ensure that all voices are heard ethically, responsibly, and safely when developing the detection models.
“FICO is impacting the lives of students, helping to build a strong representation of diversity across the data workforce for generations to come,” said Alfred Watkins, Chair of the Computer Science Department at Morehouse.
Zoldi said it’s important to get more people of color working in data science. If the people working on these problems don’t come from different backgrounds, the solutions might not be fair for everyone. Having people of color on data science teams helps find and fix biases in the data, making things more accurate and helpful across the board. It also brings new ideas and ways of thinking, which leads to better and more creative solutions.
It’s also important to get younger folks involved. People ages 18 to 35 are among the largest groups to be victims of credit card and cyber fraud. Zoldi believes that having the students at Morehouse and other HBCUs learn more about credit fraud will motivate them to prevent it.
“We try to inspire [the students] around how fraud is a really important problem for them to solve; what sort of science and math do you need to leverage, such as machine learning and behavior analytics, moreover how to make sure you can operationalize it, which is one of the most significant sorts of challenges,” Zoldi said.
According to the “2024 fraud and statistics report” from Security.org, 52 million American credit card holders have experienced credit card fraud, which has cost over $5 billion in unauthorized purchases.
Additionally, the “2024 Consumer Cyber Readiness Report” from the Cyber Civil Defense Initiative says that people of color are targeted more than their white counterparts. It also mentioned that 14 percent of Black consumers and 13 percent of Latino consumers had experienced financial loss as a result of cyber-attacks and scams.
When developing fraud detection models, Zoldi said he wants to collaborate more with HBCUs teaching data science to ensure that the scientists who create them are members of their communities. To that end, FICO plans to host the competition at other HBCUs in the spring semester.
Leave a comment