Due in class on Wednesday, March 9
In this assignment you will try to find good Bayesian network models for some
synthetic data. The data is in the files
You can copy a file with the Unix command "
cp ~ablumer/hw4a .",
where the dot can be replaced by another name if you want to use a different
name. Each file has 200 lines giving values for three binary variables, A, B,
and C, separated by tabs. See if you can find the best model using the smallest
number of parameters for each data set. Try at least A->B->C, C->B->A,
A->C<-B, and all three independent. Hand in a clear justification for
your results, with values for all your probability estimates.