Skip to content
On this page

Nature Reviews - Bayesian Statistics and Modeling

On this page

Modeling process

This is totally new to me! The big differentiator here from traditional modeling is we use preconceptions about the data to inform the model. Modeling generally has three steps:

  1. Determining the prior distribution, or beliefs coming into the experiment with beliefs as probability distributions. Usually before seeing data!

  2. Determining the likelihood function, the conditional probability of given parameters

Last updated on 3/29/2022