<< Click to Display Table of Contents >> Navigation:  Tutorials > Tutorial 10: Structure learning > Tutorial10.R

library(rSMILE)

source("License.R")

 

# Tutorial10 loads Credit10k.csv file

# and runs multiple structure learning algorithms

# using the loaded dataset.

# Use the link below to download the Credit10k.csv file:

# https://support.bayesfusion.com/docs/Examples/Learning/Credit10K.csv

 

cat("Starting Tutorial10...\n")

ds <- DataSet()

 

ds$readFile("Credit10k.csv")

cat(sprintf("Dataset has %d variables (columns) and %d records (rows)\n", 

    ds$getVariableCount(), ds$getRecordCount()))

bayesSearch <- BayesianSearch()

bayesSearch$setIterationCount(50)

bayesSearch$setRandSeed(9876543)

net1 <- bayesSearch$learn(ds)

cat(sprintf("1st Bayesian Search finished, structure score: %f\n", 

    bayesSearch$getLastScore()))

net1$writeFile("tutorial10-bs1.xdsl")

 

bayesSearch$setRandSeed(3456789)

net2 <- bayesSearch$learn(ds)

cat(sprintf("2nd Bayesian Search finished, structure score: %f\n",

    bayesSearch$getLastScore()))

net2$writeFile("tutorial10-bs2.xdsl")

 

idxAge <- ds$findVariable("Age")

idxProfession <- ds$findVariable("Profession")

idxCreditWorthiness <- ds$findVariable("CreditWorthiness")

 

backgroundKnowledge <- BkKnowledge()

backgroundKnowledge$matchData(ds)

backgroundKnowledge$addForbiddenArc(idxAge, idxCreditWorthiness)

backgroundKnowledge$addForcedArc(idxAge, idxProfession)

 

bayesSearch$setBkKnowledge(backgroundKnowledge)

net3 <- bayesSearch$learn(ds)

cat(sprintf("3rd Bayesian Search finished, structure score: %f\n",

    bayesSearch$getLastScore()))

net2$writeFile("tutorial10-bs3.xdsl")

 

tan <- TAN()

tan$setRandSeed(777999)

tan$setClassVariableId("CreditWorthiness")

net4 <- tan$learn(ds)

cat("Tree-augmented Naive Bayes finished")

net4$writeFile("tutorial10-tan.xdsl")

 

pc <- PC()

pattern <- pc$learn(ds)

net5 <- pattern$makeNetwork(ds)

cat("PC finished, proceeding to parameter learning\n")

net5$writeFile("tutorial10-pc.xdsl")

em <- EM()

matching <- ds$matchNetwork(net5)

em$setUniformizeParameters(FALSE)

em$setRandomizeParameters(FALSE)

em$setEqSampleSize(0)

em$learn(ds, net5, matching)

cat("EM finished\n")

net5$writeFile("tutorial10-pc-em.xdsl")

 

cat("Tutorial10 complete.\n")