This project demonstrates a novel approach to mitigating gender bias in machine learning models using dynamic dataset augmentation. Utilizing the Adult Income dataset, the code implements a machine learning pipeline that:
- Loads and preprocesses income prediction data
- Detects and quantifies gender bias in the dataset
- Develops a dynamic augmentation strategy to reduce bias
- Trains and evaluates models with and without bias mitigation
- Generates visualizations to compare model fairness
- Clone the repository
- Create a virtual environment (recommended):
python -m venv venv
source venv/bin/activate #Activate on windows with `venv\Scripts\activate`- Install Dependencies
pip install -r requirements.txt - Run
python main.py 