Predicting Loan Defaults Using Logistic Regression

Mentor profile
Cogitativo, Data Science Intern
Facebook, Data Science, Analytics Intern
QuantCo, Machine Learning/Quantitative Research Intern
UC Berkeley, BA
Stanford University, MS
Stanford University,
Project description

This rising senior in high school came to Polygence interested in actuary science (statistics, data modeling, and uncertainty). She had become passionate about this subject after enrolling in classes through FBLA such as Securities & Investments and Insurance & Risk Management. She had already taken AP Calc BC and AP statistics. So she came to Polygence with an eye towards making her own statistical model, allowing her to develop her data analysis and coding skills.

We will use known and unknown features pertaining to a loan candidate and the loan to predict the risk of defaulting on a loan through statistical modeling methods in R.

Predicting Loan Defaults Using Logistic Regression
Project outcome

For her final project, she used R to explore loan characteristics and how they potentially influence the probability of default. She published her research on Medium's Towards Data Science Publication.

Check out her publication
Student review

Polygence was a great experience for me. Jiying was really prepared in all of the sessions and spread the workload out evenly throughout the 10 weeks so that writing the final report wasn't too overwhelming. During each meeting, I learned about some new aspect of modeling or analyzing data that would advance the project. For homework, Jiying kept the options pretty open. I had the freedom to explore whichever part of the data I was most drawn to, and then we would discuss in the next session. Overall, I had a lot of fun and learned interesting concepts beyond the scope of a typical high school course!

Get in touch with us

Polygence logo