Given the increased flexibility of contextualized modeling, it would be helpful to encode prior knowledge to guide model learning and understand divergence from prior knowledge. It would be useful to specify priors on target parameters or the degree of heterogeneity.
Given the increased flexibility of contextualized modeling, it would be helpful to encode prior knowledge to guide model learning and understand divergence from prior knowledge. It would be useful to specify priors on target parameters or the degree of heterogeneity.