So, I’m trying to wrap my head around how to to the following: I’ve generated two prediction distributions, P1 & P2. Now I want to get a new distribution that’s basically P1/P2. In the frequentist way I would just have divided the means. But in a Bayesian fashion, I’d like to have the distribution…

Let’s say X \sim P_1 and Y \sim P_2 with the range of possible values identical between the two. Do you want to generate a distribution for X/Y or are you actually interested in the ratio of densities P_1(x) / P_2(x)? The second case is not going to be a distribution in general. It seems like you ar…

Ratio of two predictions

Questions

mattiasthalen August 24, 2020, 9:23am 3

You are correct, it was the former. Thanks!

Topic		Replies	Views
Sampling Deterministic Random Variables Questions	2	4963	March 22, 2018
How best to model risk ratio Questions	2	725	July 8, 2019
Posterior predictive distribution Questions	6	3492	May 3, 2018
Generate a predictive posterior from two RVs in time series Questions	0	375	February 18, 2021
Inferring distribution of function of random variables Questions	3	457	January 28, 2020

Ratio of two predictions

Related topics