DebiasED Tutorial

Overview

Educational data mining has transformed education by harnessing data to improve various educational dimensions. While predictive models have become core to many modern educational systems, they've also raised concerns about fairness and inherent biases in training data. This tutorial introduces participants to the DebiasED package, a comprehensive toolkit for identifying and addressing bias in educational data through pre-processing, in-processing, and post-processing techniques.

Learning Goals

By the end of this tutorial, participants will be able to:

• Identify potential types of biases in their data and understand how they propagate through models
• Apply various mitigation techniques using the DebiasED package to their own data
• Analyze the impact of bias mitigation techniques through different algorithmic fairness metrics
• Present results in terms of fairness performances

Detailed Schedule