[CS330 Chelsea Finn] Deep Multi-task learning and Meta-learning Contents

Goal: Check a higher version of the perception for robotics

Contents

1. Course introduction & start of multi-task learning 43m
2. Supervised multi-task learning, transfer learning 1h 19m
3. Meta-learning problem statement, black-box meta-learning 1h 18m
4. Optimization-based meta-learning 1h 18m
5. Few-shot learning via metric learning 1h 25m
6. Advanced meta-learning topics 1h 28m
7. Bayesian meta-learning 1h 27m
8. RL review 1h 22m
9. Multi-task RL and goal-conditioned RL 1h 25m
10. Model-based RL for multi-task learning 1h 18m
11. Meta-RL 1h 19m
    1. Meta-RL 1: Adaptable models and policies
    2. Meta-RL 2: Learning to explore
12. Offline RL and offline MTRL 1h 25m
13. Hierarchical RL and skill discovery 1h 25m
14. Lifelong learning: problem statements, forward & backward transfer 1h 25m
15. Frontiers and open problems 1h 20m