Discrete or Continuous variables? Influence or BBN on Genie? Theoretical backup
Discrete or Continuous variables? Influence or BBN on Genie? Theoretical backup
I developed a BBN of natural resources management variables with discrete variables> Could you guide me about a paper or sth for the differences and utilities of discrete_continuous variables and the BBN_influence diagramms_learning algorithms ?
Re: Discrete or Continuous variables? Influence or BBN on Genie? Theoretical backup
dear administrators please guide me, my PhD depends on genie .
-
- Site Admin
- Posts: 430
- Joined: Tue Dec 11, 2007 4:24 pm
Re: Discrete or Continuous variables? Influence or BBN on Genie? Theoretical backup
Dear Viliop,
I'm not sure I understand your question, so let me try to answer what I do understand. Discrete variables are generally approximations of continuous variables. Whenever precision is not at stake or the variable can be naturally expressed in discrete terms, I would suggest discrete variables.
As far as I know, there are no algorithms for learning influence diagrams -- all algorithms are trying to learn the world in terms of a (causal) graph expressed by a Bayesian network. Does this help?
Marek
I'm not sure I understand your question, so let me try to answer what I do understand. Discrete variables are generally approximations of continuous variables. Whenever precision is not at stake or the variable can be naturally expressed in discrete terms, I would suggest discrete variables.
As far as I know, there are no algorithms for learning influence diagrams -- all algorithms are trying to learn the world in terms of a (causal) graph expressed by a Bayesian network. Does this help?
Marek