`
小网客
  • 浏览: 1216565 次
  • 性别: Icon_minigender_1
  • 来自: 北京
社区版块
存档分类
最新评论

mahout常用距离一览表

 
阅读更多

mahout常用DistanceMeasure一览表:

org.apache.mahout.common.distance.ChebyshevDistanceMeasure;
org.apache.mahout.common.distance.CosineDistanceMeasure;
org.apache.mahout.common.distance.EuclideanDistanceMeasure;
org.apache.mahout.common.distance.MahalanobisDistanceMeasure;
org.apache.mahout.common.distance.ManhattanDistanceMeasure;
org.apache.mahout.common.distance.MinkowskiDistanceMeasure;
org.apache.mahout.common.distance.SquaredEuclideanDistanceMeasure;
org.apache.mahout.common.distance.TanimotoDistanceMeasure;
org.apache.mahout.common.distance.WeightedEuclideanDistanceMeasure;
org.apache.mahout.common.distance.WeightedManhattanDistanceMeasure;

 比较常用的有两个欧氏距离和余弦

CosineDistanceMeasure:



 EuclideanDistanceMeasure:



 EuclideanDistanceMeasure和CosineDistanceMeasure区别见图:



 其中有两个基于权重的距离计算方法:

org.apache.mahout.common.distance.WeightedEuclideanDistanceMeasure;
org.apache.mahout.common.distance.WeightedManhattanDistanceMeasure;

 使用方式如下:

WeightedManhattanDistanceMeasure cdm = new WeightedManhattanDistanceMeasure();
cdm.setWeights(new DenseVector(new double[] {1, 2, 3 }));
System.out.println(cdm.distance(new DenseVector(new double[] {0.11, 510, 10 }),
	new DenseVector(new double[] {0.11, 510, 1 })));

 设置weight即可

  • 大小: 2.5 KB
  • 大小: 1.9 KB
  • 大小: 18.6 KB
1
3
分享到:
评论

相关推荐

Global site tag (gtag.js) - Google Analytics