All Articles Education Black students receive real-world experience with FICO challenge

Black students receive real-world experience with FICO challenge

The FICO Educational Analytics Challenge helps black college students studying data science learn about AI principles in a partnership with HBCUs, writes Scott Zoldi.

5 min read

Education

Black student in front of a computer screen

(Pixabay)

Despite programs aimed at attracting more Black students to fields ranging from law to aviation, African Americans and other people of color remain underrepresented across industries. Data science is one of them; only 4.2% of data scientists are Black, even though Blacks comprise 14.4% of the US population.

In 2023, FICO created the FICO Educational Analytics Challenge that allows aspiring Black data scientists to gain experience working with real-world problems to solve in the financial services industry. In cooperation with three historically Black colleges and universities, we created an analytic problem that students could complete in about 12 weeks while learning about the principles and practices of responsible AI.

The first data science problem challenged students to identify and mitigate bias in housing datasets that factor into lenders’ mortgage decisions. We expanded the program to additional HBCUs in 2024, and this year, we challenged the students to use data science and AI techniques to detect payment card fraud.

A dynamic partnership helps students flourish

Velma Latson, PhD, an assistant professor at Bowie State University, has championed the Analytics Challenge since its inception.

“I appreciate the partnership with FICO because it has allowed the participating schools to build that relationship. It does take time,” Latson said. The Analytics Challenge will introduce a new summer data science internship program in its second year.

The human touch

The Analytics Challenge includes in-person lectures at each of the HBCUs, followed by mentoring throughout the program and beyond.

At each participating HBCU campus, I kick off the program with a keynote that lays out the significant industry challenge and starts the students on the path of how it’s addressed. Over the next 12 weeks, numerous data science team members meet regularly with students over Zoom, serving as dedicated mentors. From my perspective, this is an excellent mutual effort, partnering with students and bringing them along on an education journey. My favorite part of the program is meeting the students and having my staff mentor them.

Latson said it’s been an excellent experience for the students. “They have spent time learning the data sets, which allows them to build real-world models with their FICO mentors. It’s been invaluable for the students to learn that in the real world, data science is a messy process.”

Real-life career skills

Beyond textbooks and classroom exercises, I’ve seen how working through real-life data science challenges can give students much more confidence in their abilities and impart real-world skills.

“They are writing technical reports, and working through that process illustrates what is expected in the industry,” Latson said. “Being able to articulate in written form what you’ve learned through the data science problem so it can be easily understood helps students see the value in their English and technical writing courses. They don’t always see how these skills are applied until they’re actually needed.”

Having worked as a data scientist for over 25 years, it is a pleasure to see the students develop skills in technical writing and communication while tackling challenging data science projects.

The rewards of mentorship

The benefits of the Analytics Challenge go both ways; the program is immensely rewarding for my staff members who serve as mentors, all of whom are personally focused on giving back to the data science community.

One thing they recognize is that the students are very bright, well-prepared and professional. It’s clear that Latson and her faculty peers across the HBCUs do an excellent job preparing these students to enter the work world. Their focus is impressive, particularly since many students have jobs or family duties outside the classroom.

Furthermore, the entire FICO team is in awe of the students’ creativity and ability to engage in the data science challenge deeply. From each conversation, we learn something new. This is one of the reasons we want to expose them to the problems we work on. We learn from the students and how they look at the data.

For example, deep in the trenches of the first challenge, I was impressed by how one of the student groups was thinking about tract housing data. They pointed out that large amounts of tract housing on any given land parcel can be a proxy for race, and perhaps that data shouldn’t be used in an analytic loan originations model because it could impute bias. Other data scientists ultimately may have come to the same conclusion of imputed bias, but the eventual conversation would likely be missing that honest, passionate introspection.

Looking forward to increased diversity

Latson believes that improving diversity in data science requires an ongoing, dynamic partnership between industry and academia, a model that can easily transfer to other fields. “Our partnership with FICO is something to nurture and grow,” she said. I enthusiastically agree.

Opinions expressed by SmartBrief contributors are their own.

 


Subscribe to SmartBrief’s FREE email newsletters on social studies and civics to see the latest hot topics in education. They’re among SmartBrief’s more than 250 industry-focused newsletters.