Predicting loan defaults.

 

Since its founding in 2009, LendingClub has issued more than $22B in online loans. In this project we will explore the driving forces behind predicting good and bad loans using innovative data science technology.  

The objectives of our research include the following:

  • Identify important predictors of default risk
  • Compare performance of different prediction models 
  • Improve models by specialized tuning 
 

why is this important?

For a snapshot into the aftermath of not properly anticipating loan defaults, please watch the clip below.

 

Data Exploration

Our analysis of the LendingClub data.

Discover →

Model

How we built our predictions.

Find Out How →

Photo by alexsl/iStock / Getty Images

Results

Which loans will default?