Hello,
Can I just clarify for a report I am writing. WHat is the name for the technique that MAP uses for estimating parameter. And what what be the name for the technique that uses pm.sample (the trace) to estimate parameters.
Also am I correct in my understanding of definitions. A parameters of the gaussian process are the mean and covariance and their variable. If I set priors to these variables (i.e. the length scale) then the things that control these priors are hyperparameters. For example if I set the prior of the length scale to be an Inverse Gamma, alpha and Beta would be the hyperparameters. I’m not sure if I’m correct or not.
Thank you!