Using interpretable AI for early warning prediction in K-12 online education

• In Event: DELT- Incorporating AI in Teaching and Learning

Short Description

The purpose of this study aims to adopt interpretable AI techniques to understand key predictors identified by an early warning model. The case study analyzed data collected from 16,011 students from a K-12 online school in the US. Different interpretable AI techniques were applied to reveal overall at-risk or successful predictors, at-risk/successful types, and the learning status of individual students.

Authors

• Contributor: Jui-Long Hung, Boise State University

• Contributor: Kerry Rice, Boise State University

• Contributor: Mingyan Zhang, Zhejiang Normal University