Differential Privacy
A privacy-preserving technique that involves adding noise to data during training to prevent the recovery of sensitive information, with credit given to Apple for initiating the conversation. It's noted as becoming increasingly practical and effective.
Videos Mentioning Differential Privacy
Dawn Song: Adversarial Machine Learning and Computer Security | Lex Fridman Podcast #95
Lex Fridman
A mechanism for protecting privacy in machine learning by adding noise during the training process, providing guarantees on the inability to identify individual data points.
Ian Goodfellow: Generative Adversarial Networks (GANs) | Lex Fridman Podcast #19
Lex Fridman
A mathematical framework for quantifying privacy guarantees, enabling the design of algorithms that protect individual data.
Michael Kearns: Algorithmic Fairness, Privacy & Ethics | Lex Fridman Podcast #50
Lex Fridman
A strong and widely accepted definition of privacy that algorithms can be designed to achieve, which avoids the weaknesses of traditional anonymization methods.
Pie & AI: TensorFlow Specialization Launch @ Google HQ
DeepLearningAI
A privacy-preserving technique that involves adding noise to data during training to prevent the recovery of sensitive information, with credit given to Apple for initiating the conversation. It's noted as becoming increasingly practical and effective.