Please provide a brief summary of your paper/research.

My research explored how we can use deep learning models to estimate better the probability of default for companies based on their credit ratings. 

Traditionally, credit ratings from agencies are expressed as letter grades (AAA, AA, etc.). While helpful for categorization, these ratings are too broad to be effective for detailed financial forecasting and risk management. My study aimed to transform these letter-based ratings into numerical probabilities – essentially, measuring how likely a firm is to default given its current credit rating. 

Current models often rely on something called Gaussian copulas, which assume that both credit ratings and defaults follow a normal distribution. While simple and widely used in the industry, these models struggle to predict extreme events, and this was a shortcoming that even contributed to the 2008 financial crisis. 

To address this, I developed a deep learning-based approach that doesn’t depend on these rigid statistical assumptions. Instead, it learns directly from the data to predict risk more flexibly and accurately, especially in extreme scenarios. While the model’s results were still inconclusive due to dataset limitations, the process offered valuable experience in data handling, model design, and computational infrastructure. 

What led to your interest in this area?

I’ve been interested in finance for a long time and originally planned to build a career in that field. Later, I developed a strong passion for data science and machine learning, and this project became the perfect bridge between the two by applying cutting-edge computational methods to solve complex financial problems. 

It was an opportunity to explore how data-driven techniques could improve portfolio management, financial forecasting, and risk analysis, which are all areas that I find both intellectually challenging and practically important. 

What prompted you to submit your paper to the competition?

I received an email from Western encouraging students to submit their work to the Undergraduate Awards. At first, I wasn’t planning to apply. I didn’t think my paper had strong enough results to be competitive, since the findings were still preliminary. But a few days before the deadline, I decided to go for it anyway. It only took about half an hour to anonymize and submit the paper, and I figured there was nothing to lose. 

A few months later, I got an email saying I’d been selected as the Regional Winner for North America, which was a complete surprise! 

Can you describe how you felt when you found out you'd been selected as a winner?

I was very surprised. Since my paper didn’t have conclusive results, I expected, at best, an honorable mention. 

To find out that I’d actually won the regional category was unexpected and incredibly validating. It reminded me that even when results aren’t perfect, the research process itself – and the effort to tackle complex problems – matters a lot. 

What are your plans for the future – do you plan to continue in this field?

Right now, I’m pursuing my Master’s Degree in Statistics at Columbia University, specializing in Machine Learning. 
In the future, I hope to work as a Data Scientist or Machine Learning Engineer. I’m still open to exploring different industries, but I’m especially interested in applications of AI and data science within financial services, as well as potentially in software, health care, or marketing. 

I don’t plan to continue this specific research project, but it gave me a strong foundation in model design, data processing, and research methodology that I’ll definitely carry forward in my career.

What do you think the benefit of programs like the Undergraduate Awards is for students?

I think programs like the UA are really valuable, especially because they recognize and celebrate undergraduate research – something that often goes underappreciated. 

Getting that recognition can motivate students to keep pursuing their research interests and even publish their work in journals or present at conferences later on. 

For me personally, while the award is a wonderful recognition, the biggest benefit has been the encouragement to keep pushing forward in my research and professional development.