Imitation learning vs. offline reinforcement learning

Поделиться
HTML-код
  • Опубликовано: 12 дек 2024

Комментарии • 6

  • @TheChucknoxus
    @TheChucknoxus Год назад

    You are so good at making this stuff understandable

  • @lennartlut
    @lennartlut Год назад

    Fascinating and well presented. Thank you!

  • @alexanderchernyavskiy9538
    @alexanderchernyavskiy9538 2 года назад +7

    an excellent talk! but it would be nice if it was a bit louder

  • @youness7230
    @youness7230 2 года назад

    Excellent talk, thank you sir.

  • @randywelt8210
    @randywelt8210 2 года назад +1

    From BC supervised RL, to online/offline.. RL has still a supervised reward structure (sensor feedback), where genetic are the least supervised random case. Anyways, the RL feedback loop is always true for all living creatures on earth. I would regard RL as the super term on a continuous supervision spectrum dependent on the skill level of sensor feedback.

  • @prof_shixo
    @prof_shixo 2 года назад +1

    Nice lecture, but the sound is very low. Tyr to use a better microphone next time.