Enhancing AI Fairness: Leveraging Differential Privacy

Explore how outliers in time series prediction models introduce bias and learn how differential privacy can mitigate these issues in this 47-minute conference talk from RSA Conference. Saahil Shenoy PhD, Founding AI Scientist at Bedrock Security, demonstrates a novel approach to using differential privacy for outlier removal that improves both model robustness and privacy. Learn practical steps to enhance AI/ML model fairness and reliability through this innovative methodology that combines privacy-preserving techniques with improved model performance.