Fairness in Healthcare ML: Beyond Accuracy Metrics

In healthcare ML, overall accuracy is not sufficient.

Models must be evaluated for fairness across different populations.

Challenges include:

• Imbalanced datasets
• Underrepresentation of certain groups
• Bias in data collection

Key practices:

• Subgroup performance analysis
• Bias detection methods
• Continuous monitoring

Fairness must be integrated into the development and deployment process.

My work focuses on applying ML with this broader perspective.

I am open to remote roles globally.