How are neural networks distinct than Bayesian belief

Question

How are neural networks distinct than Bayesian belief networks all along the following dimensions:

(A) Inspect capability of knowledge,
(B) Requirement for probabilities obtained from "domain" experts,
(C) Requirement for data to train the system, and
(D) Capability of the system to make categorization based on input data.

Answer 1

Bayesian belief networks are inspectable, known probabilities are needed, training data are not required, and they can categorize into multiple categories.

Neural networks are not inspectable, they do not require domain expertise or known probabilities, training data are needed, and they are best for binary classification ("no" or "yes").