I work with JS programmer on the logic for a web app. We have factors that influence a composition of a set.
Envision rows in Excel that tell a set to have 5 members or 10, etc. Each row has certain "properties" or "weights" such as "If this row is implemented then the set will hold 10 units and nothing else" or if row two is implemented "then the set will hold 5 units and whatever row 8 says." Like rules on each row that determine the set.
My question: This can be done by a human inductively by going through each row and just doing it but is the logic better suited for setting up a Bayesian Network with SPSS or Bayes Server or is should I be looking at multi-variate analysis.