About Our Model
Expanding from prior neural network models to predict Coronary Heart Disease, our project leverages adversarial debiaising to mitigate gender bias.
Development
Our development process began with baseline models, followed by the primary neural network. From there, we implemented an adversarial module, and concluded with hyperparameter tuning.
Results & Beyond
Analysis of our model’s performance in both accuracy and fairness. Discussion on limitations, implications, and potential avenues of further research.