data set.
JID=ID(Interval) , x7_1= age group(1=old,0=young) , AY~AB is disease (1=infection,0=uninfection) (x7_1,AY~AB is BINARY)
i want to know the association of variables which is disease name using the hpbnet of EM.
can I know a network in diseases using bayesian network node?
But I know nothing from seting data(ex. what is the TARGET variable?) and what should I choose which bayesian network.
SO, what I want is first, which bayesian network (NAIVE, BAN, TAN) shoud I use?
secend, if I wanna know network in disease, how can I set the data?(significance level, network model, maximum parents, number of bins)
The Bayesian Network implementation in EM is a classifier. You can think of it as an alternative to models like decision trees, neural networks, etc. that makes certain simplifying assumptions about the relationships between the inputs and target (the variable containing the event or class you want to predict).
It it is not clear that you have a unique target variable. I would suggest getting more familar with HPBNET, starting with video below. That will help you decide whether it is actually appropriate for your situation. (Off the top of my head I would consider clustering the diseases or computing the distances between the diseases and projecting the distance matrix onto two dimensions. But there are other possibilities and I'm not entirely sure of the desired end result)
i hope this helps,
Ray
Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.
Register today!Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.
Find more tutorials on the SAS Users YouTube channel.