SMLG (Statistical Machine Learning Group) Discussion Forum

The entropy of two different parent sets giving 2 variables are the following. The following is for sprinkler data set.

Suppose X = Fertilizer, Y = MyLawnGreen, W = Sprinkler , Z = NBLawnGreen, W ' = NBLawnWet, Z ' = MyLawnWet

p-value of ( X ind Y | W, Z) = 0.8169111, p-value of ( X ind Y | W ' , Z ' ) = 0.01652227

log Score ( X , Pa (x) = { W, Z } = -52.7119 , log Score ( X , Pa (x) = { W ' , Z ' } = -41.405

entrop (Nwz) = 1.241856 entrop (Nw'z') = 1.204842

The attachment is the result of Conditional Test Result for sprinkler dataset.

This attachment is Conditional Independence Test for variable pairs giving one parent.

Suppose X = Fertilizer, Y = MyLawnGreen, W = Sprinkler , Z = NBLawnGreen, W ' = NBLawnWet, Z ' = MyLawnWet

log Score ( Y , Pa (x) = { W, Z } )= -71.864 , log Score ( X , Pa (x) = { W ' , Z ' } )= -58.1877

entrop (Nxwz) = 1.667553 entrop (Nxw'z') = 1.519601

entrop (Nywz) = 1.841074 entrop (Nyw'z') = 1.678443

The following is the result of independence Test.
For example:

79 1, 5 | 0, 6 3.834246e-07

The first column is the row number. The 2nd column means children are 1, 5 and parents are 0,6. The 3rd column means the p-value of children 1,5 giving parents 0, 6.

The following is the entropy of the dataset.

For example:

51 3, 5, 7 1.773572

The first column is the row number. The 2nd column means variables are 3, 5, 7. The 3rd column means the entropy of the variable sets is 1.773572.

The following is sorted entropy.

The attachment is Conditional Independence Test result which is sorted based on p-value.

The attached is the conditional independence test and p-value for sprinkler data sets with 1000 samples.

The attached is the entropy for sprinkler data sets with 1000 samples.