Imagine you have a sequence of models, for example each one being an equation (or a system of equations): they are connected to each other so that the outputs of a model are used as inputs for one or more other models.
Is there a specific terminology to call this structure? I am trying to find literature on the subject, but I realized I am probably missing some keywords. I tried with "model chains", "model networks", and similar names, but I don't feel that's the right nomenclature.
I think there is a specific terminology for sub-categories of this structure: for example, Bayesian networks could be considered network of models, each node being a probabilistic model described by a conditional probability table. But what if the model inside one of the nodes was deterministic? How would you call the structure, then?
Sorry if the question is naive; in the beginning I thought it would be easy to find an answer, but I found myself skimming through tons of literature without finding anything promising.
Thank you in advance for any help you can provide!
Dear Alberto, I also think that terminology is very important, and each word has its own meaning that must be fully understood in the context of what we intend to express. So, in my opinion, your question only shows your intelligence. Dynamic Programming (DP) is a very powerful technique and has an interesting terminology for sequential decisions problems in discrete optimization; perhaps you can adapt the terminology (or eventually adopt it) to your problem. Sometimes, the technique may be not easy to understand fully or may be not easy modeling the problem, but the terminology is quite intuitive and easy of understanding. I do not know if this may help, but because you spoke about (or I understood) sequential decision problems, I remembered immediately DP.
- Badges
- Science topic
- Similar topics
- Philosophy
- Esthetics
- Interpretation