search

Richie-Lee / Bayesian_AB_Test

MSc_Thesis code repository
0 stars 0 forks source link

Prior odds - how to implement in inference #7

Open Richie-Lee opened 9 months ago

Richie-Lee commented 9 months ago

For the post odds, it's implementation is quite straight-forward.

However, when considering metrics such as uplift, Prob-beat-control and loss. How do we incorporate those? By simply updating beta functions, it only uses the priors inside the marginal likelihoods

Richie-Lee commented 9 months ago

Arguably for simplicity, it would be reasonably to just take post odds of 1