Decision network

• Do you k开发者_如何转开发now what is this, and How they represent decisions (using probabilities) in Decision Theory
  • Are they similar to decision trees?

---

Yes, they share similar concepts, and both mainly contains: decision nodes & chance nodes.

Except in a network, you don't necessarily start from one node but it could start with a combination of different nodes, and instead of every node taking a single input in tree, in a network, a decision/chance could be based on multiple inputs.

e.g. in a single node in a network : Weather=Sunny + Holiday=True => go for a hike=True. so you dont need to split it into multiple nodes as in trees.

In general, their differences are similar as to trees and graphs in a traditional sense. You should be able to work out a decision network easily if you are familiar with decision trees.

Hope this is helpful.