决策树的应用

import org.apache.spark.mllib.linalg.Vectors
import org.apache.spark.mllib.regression.LabeledPoint
import org.apache.spark.mllib.tree.DecisionTree
import org.apache.spark.mllib.util.MLUtils
import org.apache.spark.{SparkConf, SparkContext}

object decisionT{
  val conf=new SparkConf().setMaster("local").setAppName("DecisionTree")
  val sc=new SparkContext(conf)
  val format=Map[Int,Int]()
  def main(args: Array[String]): Unit = {
    val source=MLUtils.loadLibSVMFile(sc,"file:\\c:\\DATA\\sample_libsvm_data.txt")
    val model = DecisionTree.trainClassifier(source,2,format,"entropy",5,3)
    //Vectors.dense(Array(124,124,265,352,316,351,363))
    //val st="1 159:121 160:254 161:136 186:13 187:230 188:253 189:248 "
    val testTree=LabeledPoint(1, Vectors.dense(121,254,136 ,113 ,20 ))
    println(model.predict(testTree.features))
  }


}

建立的测试数据集如下：
0 1:51 2:159 3:253 4:159 5:50  
1 1:124 2:253 3:255 4:63 5:96
1 1:145 2:255 3:211 4:31 5:32 
0 1:64 2:253 3:255 4:63 5:96 
1 1:121 2:254 3:136 4:13 5:230  

运行预测的结果如下：

18/01/25 11:44:32 INFO BlockManager: Removing RDD 14
1.0
18/01/25 11:44:32 INFO RandomForest: Internal timing for DecisionTree:
18/01/25 11:44:32 INFO RandomForest:   init: 1.855486899
  total: 2.098156533
  findSplitsBins: 1.613379893
  findBestSplits: 0.229065344
  chooseSplits: 0.221569245