-
Notifications
You must be signed in to change notification settings - Fork 0
/
TuningDagging.java
62 lines (46 loc) · 2.11 KB
/
TuningDagging.java
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
import weka.classifiers.Evaluation;
import weka.classifiers.meta.CVParameterSelection;
import weka.classifiers.meta.Dagging;
import weka.classifiers.meta.LogitBoost;
import weka.core.Instances;
import weka.core.Utils;
import java.io.BufferedReader;
import java.io.FileReader;
import java.io.FileWriter;
import java.io.PrintWriter;
public class TuningDagging {
public static void main(String[] args) throws Exception{
// load data sets
Instances train = new Instances(
new BufferedReader(
new FileReader("/Weka-3-6/ProjectMilestone5/mfeat-factors_train.arff")));
Instances test = new Instances(
new BufferedReader(
new FileReader("/Weka-3-6/ProjectMilestone5/mfeat-factors_test.arff")));
train.setClassIndex(train.numAttributes() - 1);
test.setClassIndex(test.numAttributes()-1);
CVParameterSelection ps = new CVParameterSelection();
ps.setClassifier(new Dagging());
//find optimal parameter
ps.setOptions(weka.core.Utils.splitOptions("-P \"F 1.0 10.0 1.0\" -P \"S 1.0 10.0 1.0\" -W \"weka.classifiers.meta.Dagging\" -- -W \"weka.classifiers.trees.LMT\" -- -I 5 -A"));
Dagging cls = new Dagging();
//change the base classifier
ps.buildClassifier(train);
System.out.println(Utils.joinOptions(ps.getBestClassifierOptions()));
String[] options=ps.getBestClassifierOptions();
//change the parameter for dagging
cls.setOptions(options);
cls.buildClassifier(train);
PrintWriter pw=new PrintWriter(new FileWriter("/Weka-3-6/ProjectMilestone5/mfeat-factors-L5.txt"));
for (int i = 0; i < test.numInstances(); i++) {
double pred = cls.classifyInstance(test.instance(i));
pw.println(pred);
}
pw.close();
//weka.core.SerializationHelper.write("/Weka-3-6/ProjectMilestone2/hypothyroid21.model", cls);
Evaluation eval=new Evaluation(train);
eval.evaluateModel(cls,test);
Double error_c=eval.errorRate();
System.out.println(error_c);
}
}