Rule-based Bayesian regression

We introduce a novel rule-based approach for handling regression problems. The new methodology carries elements from two frameworks: (i) it provides information about the uncertainty of parameters interest using Bayesian inference, and (ii) allows incorporation expert knowledge through systems. blending those different frameworks can be particularly beneficial various domains (e.g., engineering), where even though significance quantification motivates approach, there is no simple way to incorporate researcher intuition into model. validate our models by applying them synthetic applications: linear problem more complex structures based on partial differential equations, we illustrate their use cases derived real data. Finally, review advantages methodology, which include simplicity implementation, reduction due added and, in some occasions, derivation better point predictions, outline limitations, mainly computational complexity perspective, such as difficulty choosing an appropriate algorithm burden.