Sign inSign in

Sergey Levine

Professor of Electrical Engineering and Computer Sciences · UC Berkeley · Distinguished

Sergey Levine is a Professor of Electrical Engineering and Computer Sciences at UC Berkeley who works on reinforcement learning and robotics and teaches CS 285.

More

Sergey Levine is a Professor of Electrical Engineering and Computer Sciences at UC Berkeley who works on reinforcement learning and robotics and teaches CS 285. His graduate lectures derive REINFORCE with baselines, analyse covariate shift and DAgger, and build actor-critic methods, TRPO and variational inference.

lectures
69
series
1
hours
23

Start here

Series and courses

UC Berkeley CS 285: Deep Reinforcement Learningfrom Sutton & Barto, Reinforcement Learning: An Introduction69 lectures · 23 h
Open the course
  1. 1Foundations of Deep Reinforcement Learning: Problem Formulation & Neural Control17:55
  2. 2Sequential Decision-Making: Policy Gradients vs. Value-Based Architectures29:02
  3. 3Imitation Learning I: Behavioral Cloning & Distributional Shift24:48
  4. 4Imitation Learning II: Covariate Shift & Dataset Aggregation (DAgger)23:06
Show 65 more
  1. 5Imitation Learning III: Inverse RL & Generative Adversarial Imitation Learning (GAIL)32:23
  2. 6Policy Gradient Foundations: Trajectory Distributions & High-Dimensional State Spaces26:29
  3. 7Deep Reinforcement Learning9:11
  4. 8Policy Gradients I: Deriving the REINFORCE Estimator & Policy Objective14:02
  5. 9Exploration of fundamental concepts in deep reinforcement learning13:17
  6. 10Policy Gradients II: Variance Reduction, Baselines & Reward-to-Go Causality14:52
  7. 11Actor-Critic Methods: Value Function Bootstrapping & Advantage Estimation15:41
  8. 12Exploration of deep reinforcement learning architectures and algorithms13:32
  9. 13Natural Policy Gradients, Covariant Updates & The Fisher Information Matrix25:13
  10. 14Trust Region Policy Optimization (TRPO) & Monotonic Policy Improvement17:29
  11. 15Exploration of fundamental concepts in deep reinforcement learning18:32
  12. 16Exploration of advanced topics in deep reinforcement learning15:55
  13. 17Foundational concepts in deep reinforcement learning16:17
  14. 18Exploration of fundamental concepts in deep reinforcement learning15:08
  15. 19Exploration of deep reinforcement learning architectures and algorithms17:02
  16. 20Foundational concepts in deep reinforcement learning, likely13:38
  17. 21Exploration of deep reinforcement learning techniques11:57
  18. 22Deep Reinforcement Learning23:41
  19. 23Exploration of advanced topics in deep reinforcement learning10:04
  20. 24Exploration of advanced topics in deep reinforcement learning11:00
  21. 25Deep Reinforcement Learning21:23
  22. 26Exploration of deep reinforcement learning, likely18:48
  23. 27Exploration of advanced topics in deep reinforcement learning21:08
  24. 28Foundational concepts in deep reinforcement learning19:20
  25. 29Exploration of advanced topics in deep reinforcement learning23:12
  26. 30Exploration of advanced topics in deep reinforcement learning23:49
  27. 31Exploration of advanced topics in deep reinforcement learning13:23
  28. 32Fundamental concepts in deep reinforcement learning17:47
  29. 33Exploration of advanced topics in deep reinforcement learning9:35
  30. 34Exploration of deep reinforcement learning algorithms17:09
  31. 35Exploration of advanced topics in deep reinforcement learning17:22
  32. 36Model-Based Reinforcement Learning with Policies15:06
  33. 37Model-Based Reinforcement Learning with Policies15:04
  34. 38Model-Based Reinforcement Learning with Policies29:20
  35. 39Foundational concepts in deep reinforcement learning19:52
  36. 40Exploration of advanced topics in deep reinforcement learning15:49
  37. 41Exploration of advanced topics in deep reinforcement learning14:31
  38. 42Deep Reinforcement Learning14:17
  39. 43Exploration of advanced topics in deep reinforcement learning15:28
  40. 44Exploration of advanced topics in deep reinforcement learning13:52
  41. 45Offline Reinforcement Learning38:01
  42. 46Offline Reinforcement Learning25:33
  43. 47Offline Reinforcement Learning (CS 285, Lecture 15, Part 3)21:20
  44. 48Offline Reinforcement Learning II31:59
  45. 49Offline Reinforcement Learning II18:17
  46. 50RL Theory50:10
  47. 51RL Theory21:59
  48. 52Variational Inference, Part 1 (CS 285, Lecture 18)20:13
  49. 53Variational Inference, Part 219:37
  50. 54Variational Inference, Part 317:52
  51. 55Variational Inference, Part 4 (CS 285, Lecture 18)25:29
  52. 56Control as Inference, Part 121:06
  53. 57Control as Inference, Part 228:39
  54. 58Control as Inference, Part 321:26
  55. 59Control as Inference, Part 510:58
  56. 60Inverse Reinforcement Learning, Part 123:50
  57. 61Inverse Reinforcement Learning, Part 212:29
  58. 62Inverse Reinforcement Learning, Part 414:22
  59. 65RL with Sequence Models & Language Models, Part 1 (CS 285, Lecture 21)29:54
  60. 66RL with Sequence Models & Language Models, Part 223:40
  61. 67RL with Sequence Models & Language Models, Part 316:59
  62. 68Transfer Learning & Meta-Learning44:17
  63. 69Transfer Learning & Meta-Learning (CS 285, Lecture 22, Part 5)13:26
  64. 70Challenges and Open Problems in Deep Reinforcement Learning (CS 285, Lecture 23, Part 1)28:25
  65. 71Challenges & Open Problems in Deep Reinforcement Learning38:01

In these lectures

Draws on

People cited in the lectures, at the moment of the citation.

  • Daniel WolpertNeuroscientist working on motor control, cited for his quote on brains and movement.7:38
  • Richard BellmanPioneer of Markov decision processes in the 1950s7:50
  • Bert KappenWork on KL-divergence control5:45
  • Richard SuttonPioneer of reinforcement learning in computer science.2:41
  • Yael NivPsychology Department & Princeton Neuroscience Institute, Princeton University, cited for reinforcement learning in the brain26:29
  • Alan TuringQuoted on simulating the child's mind versus the adult mind28:42

About

Sergey Levine is a Professor of Electrical Engineering and Computer Sciences at UC Berkeley who works on reinforcement learning and robotics and teaches CS 285. His graduate lectures derive REINFORCE with baselines, analyse covariate shift and DAgger, and build actor-critic methods, TRPO and variational inference.

Connections

75

Their places, courses, books and lectures.

Institutions3

Concept1

Lectures69

Show 63 more