Leveraging-Explainable-AI-to-Mitigate-Bias

Team Lead: Rowan White
Participants: Aaron Barthwal, Khiem Nguyen, Laya Srinivas, Ram Gudur, Sanya Oak

Poster

Introduction

Artificial intelligence is rapidly gaining traction in the world of technology as a powerful tool for data analysis, classification, and prediction. However, today’s AI typically relies on a ”black box” model: a human observer can’t extrapolate how the model makes its decisions through its operations alone. We rely on quantifying the model through its output. Explainable AI (XAI) is a hot topic in AI research today, endeavoring to make the inner workings of a ”black box” model clearer by interpreting the results

Dataset

Our Model

SHAP and LIME

Implementing Bias

Results