|
<< Click to Display Table of Contents >> Navigation: Tutorials > Tutorial 10: Structure learning > Tutorial10.cs |
using System;
using Smile;
using Smile.Learning;
// 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
namespace SmileNetTutorial
{
class Tutorial10
{
public static void Run()
{
Console.WriteLine("Starting Tutorial10...");
DataSet ds = new DataSet();
try
{
ds.ReadFile("Credit10k.csv");
} catch (SmileException)
{
Console.WriteLine("Dataset load failed");
return;
}
Console.WriteLine("Dataset has {0} variables (columns) and {1} records (rows)",
ds.VariableCount, ds.RecordCount);
BayesianSearch bayesSearch = new BayesianSearch();
bayesSearch.IterationCount = 50;
bayesSearch.RandSeed = 9876543;
Network net1;
try
{
net1 = bayesSearch.Learn(ds);
} catch (SmileException)
{
Console.WriteLine("Bayesian Search failed");
return;
}
Console.WriteLine("1st Bayesian Search finished, structure score: {0}",
bayesSearch.LastScore);
net1.WriteFile("tutorial10-bs1.xdsl");
Network net2;
bayesSearch.RandSeed = 3456789;
try
{
net2 = bayesSearch.Learn(ds);
}
catch (SmileException)
{
Console.WriteLine("Bayesian Search failed");
return;
}
Console.WriteLine("2nd Bayesian Search finished, structure score: {0}",
bayesSearch.LastScore);
net2.WriteFile("tutorial10-bs2.xdsl");
int idxAge = ds.FindVariable("Age");
int idxProfession = ds.FindVariable("Profession");
int idxCreditWorthiness = ds.FindVariable("CreditWorthiness");
if (idxAge < 0 || idxProfession < 0 || idxCreditWorthiness < 0)
{
Console.WriteLine("Can't find dataset variables for background knowledge");
Console.WriteLine("The loaded file may not be Credit10k.csv");
return;
}
BkKnowledge backgroundKnowledge = new BkKnowledge();
backgroundKnowledge.MatchData(ds);
backgroundKnowledge.AddForbiddenArc(idxAge, idxCreditWorthiness);
backgroundKnowledge.AddForcedArc(idxAge, idxProfession);
bayesSearch.BkKnowledge = backgroundKnowledge;
Network net3;
try
{
net3 = bayesSearch.Learn(ds);
}
catch (SmileException)
{
Console.WriteLine("Bayesian Search failed");
return;
}
Console.WriteLine("3rd Bayesian Search finished, structure score: {0}",
bayesSearch.LastScore);
net3.WriteFile("tutorial10-bs3.xdsl");
Network net4;
TAN tan = new TAN();
tan.RandSeed = 777999;
tan.ClassVariableId = "CreditWorthiness";
try
{
net4 = tan.Learn(ds);
}
catch (SmileException)
{
Console.WriteLine("TAN failed");
return;
}
Console.WriteLine("Tree-augmented Naive Bayes finished");
net4.WriteFile("tutorial10-tan.xdsl");
PC pc = new PC();
Pattern pattern;
try
{
pattern = pc.Learn(ds);
} catch (SmileException)
{
Console.WriteLine("PC failed");
return;
}
Network net5 = pattern.MakeNetwork(ds);
Console.WriteLine("PC finished, proceeding to parameter learning");
net5.WriteFile("tutorial10-pc.xdsl");
EM em = new EM();
DataMatch[] matching;
try
{
matching = ds.MatchNetwork(net5);
}
catch (SmileException)
{
Console.WriteLine("Can't automatically match network with dataset");
return;
}
em.UniformizeParameters = false;
em.RandomizeParameters = false;
em.EqSampleSize = 0;
try
{
em.Learn(ds, net5, matching);
}
catch (SmileException)
{
Console.WriteLine("EM failed");
return;
}
Console.WriteLine("EM finished");
net5.WriteFile("tutorial10-pc-em.xdsl");
Console.WriteLine("Tutorial10 complete");
}
}
}