This is just Bayesian modeling. Move the parameters into the model, push the hyperparameters into a prior. Move the hyperparameters into the model, push the hyp

This is just Bayesian modeling. Move the parameters into the model, push the hyperparameters into a prior. Move the hyperparameters into the model, push the hyperhyperparameters into a hyperprior, etc. Choosing hyperparameters from minor tuning or literature review or domain expertise or whatever is just a form of creating a prior.