[Case Study]Decision Tree Classification using Java

是 否 +2 论坛币

k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群

赵安豆老师微信：zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

立即领取

感谢您参与论坛问题回答

经管之家送您两个论坛币！

+2 论坛币

1. */
   
2. // scalastyle:off println
   
3. package org.apache.spark.examples.ml;
   
4. // $example on$
   
5. import org.apache.spark.SparkConf;
   
6. import org.apache.spark.api.java.JavaSparkContext;
   
7. import org.apache.spark.ml.Pipeline;
   
8. import org.apache.spark.ml.PipelineModel;
   
9. import org.apache.spark.ml.PipelineStage;
   
10. import org.apache.spark.ml.classification.DecisionTreeClassifier;
    
11. import org.apache.spark.ml.classification.DecisionTreeClassificationModel;
    
12. import org.apache.spark.ml.evaluation.MulticlassClassificationEvaluator;
    
13. import org.apache.spark.ml.feature.*;
    
14. import org.apache.spark.sql.DataFrame;
    
15. import org.apache.spark.sql.SQLContext;
    
16. // $example off$
    
17. 
    
18. public class JavaDecisionTreeClassificationExample {
    
19.   public static void main(String[] args) {
    
20.     SparkConf conf = new SparkConf().setAppName("JavaDecisionTreeClassificationExample");
    
21.     JavaSparkContext jsc = new JavaSparkContext(conf);
    
22.     SQLContext sqlContext = new SQLContext(jsc);
    
23. 
    
24.     // $example on$
    
25.     // Load the data stored in LIBSVM format as a DataFrame.
    
26.     DataFrame data = sqlContext.read().format("libsvm").load("data/mllib/sample_libsvm_data.txt");
    
27. 
    
28.     // Index labels, adding metadata to the label column.
    
29.     // Fit on whole dataset to include all labels in index.
    
30.     StringIndexerModel labelIndexer = new StringIndexer()
    
31.       .setInputCol("label")
    
32.       .setOutputCol("indexedLabel")
    
33.       .fit(data);
    
34. 
    
35.     // Automatically identify categorical features, and index them.
    
36.     VectorIndexerModel featureIndexer = new VectorIndexer()
    
37.       .setInputCol("features")
    
38.       .setOutputCol("indexedFeatures")
    
39.       .setMaxCategories(4) // features with > 4 distinct values are treated as continuous
    
40.       .fit(data);
    
41. 
    
42.     // Split the data into training and test sets (30% held out for testing)
    
43.     DataFrame[] splits = data.randomSplit(new double[]{0.7, 0.3});
    
44.     DataFrame trainingData = splits[0];
    
45.     DataFrame testData = splits[1];
    
46. 
    
47.     // Train a DecisionTree model.
    
48.     DecisionTreeClassifier dt = new DecisionTreeClassifier()
    
49.       .setLabelCol("indexedLabel")
    
50.       .setFeaturesCol("indexedFeatures");
    
51. 
    
52.     // Convert indexed labels back to original labels.
    
53.     IndexToString labelConverter = new IndexToString()
    
54.       .setInputCol("prediction")
    
55.       .setOutputCol("predictedLabel")
    
56.       .setLabels(labelIndexer.labels());
    
57. 
    
58.     // Chain indexers and tree in a Pipeline
    
59.     Pipeline pipeline = new Pipeline()
    
60.       .setStages(new PipelineStage[]{labelIndexer, featureIndexer, dt, labelConverter});
    
61. 
    
62.     // Train model.  This also runs the indexers.
    
63.     PipelineModel model = pipeline.fit(trainingData);
    
64. 
    
65.     // Make predictions.
    
66.     DataFrame predictions = model.transform(testData);
    
67. 
    
68.     // Select example rows to display.
    
69.     predictions.select("predictedLabel", "label", "features").show(5);
    
70. 
    
71.     // Select (prediction, true label) and compute test error
    
72.     MulticlassClassificationEvaluator evaluator = new MulticlassClassificationEvaluator()
    
73.       .setLabelCol("indexedLabel")
    
74.       .setPredictionCol("prediction")
    
75.       .setMetricName("precision");
    
76.     double accuracy = evaluator.evaluate(predictions);
    
77.     System.out.println("Test Error = " + (1.0 - accuracy));
    
78. 
    
79.     DecisionTreeClassificationModel treeModel =
    
80.       (DecisionTreeClassificationModel) (model.stages()[2]);
    
81.     System.out.println("Learned classification tree model:\n" + treeModel.toDebugString());
    
82.     // $example off$
    
83.   }
    
84. }

复制代码

扫码加我 拉你入群

请注明：姓名-公司-职位

以便审核进群资格，未注明则拒绝

分享0 收藏0 回帖

关键词：Case study Decision cation study ATION example import Java