I can think of:

  • Factorial
  • Hierarchical
  • Layered
  • Nested
  • 'standard'

enter image description here



I think your picture coming from Murphy's tutorial on graphical model.

And all these models can be named as "directed probabilistic graphical model" or "dynamic Bayesian network".

There are two problems we want to consider on these types of structured models

  • What is the structure (connectivity among random variables)
  • What are the hidden variables (which random variable we cannot observe.)

One can use knowledge specify the structure or learn the structure from data, Examples can be found here. Which variables are hidden usually comes from domain knowledge.

In sum, the name of the model does not matter too much. All of them are directed probabilistic graphical models.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.