1
$\begingroup$

As I understand, Decision graph is a directed graph where each vertex is a "Question" (decision to make), and each edge is an "Answer" (decision made for the vertex this edge is coming from). The terminal nodes of that graph is a solution.

Problem: With this structure - each "Answer" can lead only to a single consequent "Question". However in my real-world problem "Answer" can lead to multiple consequent "Questions".

What is the right data structure and algorithms for such kind of decision making , finding optimal solution problems.

$\endgroup$
4
  • $\begingroup$ It's still a decision tree. You can put several questions at one vertex. $\endgroup$ Mar 2, 2017 at 1:13
  • $\begingroup$ Not exactly. Imagine you have 2 consequent questions with 3 answers each. How will you fit it in one node? How many resolutions you will have? $\endgroup$ Mar 5, 2017 at 17:25
  • $\begingroup$ You will have $3^2=9$ resolutions. $\endgroup$ Mar 5, 2017 at 18:43
  • $\begingroup$ Exactly - that will be a Cartesian product of all sets of answers on the same level, so it's not very nice. Is there any better data structure/algorithm for that? $\endgroup$ Mar 9, 2017 at 14:26

0

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Browse other questions tagged or ask your own question.