topic Re: interpretation bayesian networks in SAS Data Science

interpretation bayesian networks

PICKME — Thu, 25 Aug 2016 12:06:15 GMT

FIRST, data set

x7_1 : 1(old) 0(young)

AA,AB, ~ , DN,DM : 1 (disease o) 0(disease x)

JID : person

SECOND, i analyzed bayesian network node in SAS E Minser.

ID :JID

TARGET : x7_1

INPUT : AA,AB, ~ , DN,DM

network model: TAN

number of bins: 2

maximum parents: 1

THIRD, analysis results

Why is the result like that?

The result what i want is this.

-> x7_1 1 CV 0 probability

x7_1 1 CV 1 probability

x7_1 0 CV 0 probability

x7_1 0 CV 1 probability

could you plz help me..

JasonXin — Thu, 25 Aug 2016 15:45:25 GMT

Hi, this is from Enterprise Miner? Jason Xin

PICKME — Thu, 25 Aug 2016 16:16:49 GMT

YES!!!!!

PLZ. I'M SO CURIOUS!!!!!

yeliu — Thu, 25 Aug 2016 17:28:59 GMT

In HP BN Classifier node, if you specify network model to be TAN, the maxparents are not applied.

If you really want a TAN model with maxparents=1, you need to spcify Naive Bayes in your network model.