On the sensitivity of the marginal likelihood to prior choice

Поділитися
Вставка
  • Опубліковано 25 лис 2024

КОМЕНТАРІ • 5

  • @stevenreyes1284
    @stevenreyes1284 5 років тому +1

    This was a really nice illustration. Thanks for all of your videos, they're very helpful!

  • @timurgareev6356
    @timurgareev6356 5 років тому

    Hi, Ben! Where can we find the notebook that you use in the video to play around?

  • @janiceliu5473
    @janiceliu5473 2 роки тому

    I don't really understand why Marginal Likelihood (P(X)) change with prior (mu), since it's already integrated over mu. Shouldn't it be a constant?

    • @janiceliu5473
      @janiceliu5473 2 роки тому

      I think I'd like to give some clarification here after I plot the change of the marginal likelihood with the prior in python. The plots in the video are not very clear. The prior should be a function of a but not mu. The same for the marginal likelihood, it's also not a function of mu, as mu is integrated out. On the contrary, likelihood and posterior are functions of mu.