Using Statistics and Machine Learning to Predict Student Success

Presentation Summary

Most higher education institutions have implemented student success initiatives towards helping students succeed through their college career. Higher education institutions also are faced with limited resources for implementing student success initiatives. At Purdue, we are using statistical modelling and machine learning techniques to predict student outcomes before they happen, allowing Purdue to target student success resources at students that are most likely to perform poorly. Our presentation will focus on the methodology we used for building our models, the features of the data which ended up being most important to student success, how we plan to continue improving our models, and finally how Purdue is operationalizing these learnings.

Presentation Speaker(s)

Monal Patel
Ian Pytlarz

Presentation Files

You must be a registered member of the HEDW and properly logged in to view this content