Doctoral studies

After completing the course, students will be able to critically analyze and apply methods for ensuring privacy (e.g., differential privacy, federated learning), fairness, and explainability in machine learning systems, while understanding the implications of global privacy laws such as GDPR and CCPA. They will also be able to evaluate challenges related to AI alignment, including specification gaming and reward hacking, and design AI systems that are ethically sound and socially responsible.

Content:

• Privacy and Data Protection: Privacy-preserving machine learning (e.g., differential privacy, federated learning), global privacy laws (GDPR, CCPA, etc.)
  • Fairness in machine learning and AI, measures for its evaluation and techniques for improving it
  • Explainable and interpretable AI, measures for its evaluation and techniques for improving it
  • The AI alignment problem, algorithmic games, specification gaming and reward hacking