距离和相似度测量
距离或相似度的不同测量方法便于不同类型的分析. Wolfram 语言提供内置函数用于许多标准的距离测量,以及给出任意测量的符号定义.
数值数据
EuclideanDistance ▪ SquaredEuclideanDistance ▪ NormalizedSquaredEuclideanDistance ▪ ManhattanDistance ▪ ChessboardDistance ▪ BrayCurtisDistance ▪ CanberraDistance ▪ CosineDistance ▪ CorrelationDistance ▪ BinaryDistance ▪ WarpingDistance ▪ CanonicalWarpingDistance
布尔数据
HammingDistance ▪ JaccardDissimilarity ▪ MatchingDissimilarity ▪ DiceDissimilarity ▪ RogersTanimotoDissimilarity ▪ RussellRaoDissimilarity ▪ SokalSneathDissimilarity ▪ YuleDissimilarity
字符串数据
EditDistance ▪ DamerauLevenshteinDistance ▪ HammingDistance ▪ SmithWatermanSimilarity ▪ NeedlemanWunschSimilarity
图像与颜色
地理空间与时间数据
GeoDistance ▪ TravelDistance ▪ TravelTime ▪ DateDifference
通用与混合数据
FeatureDistance — 通过数据训练所得特征空间中的距离
DistanceFunction — 指定距离函数的选项
DistanceMatrix — 元素间成双距离的矩阵
Nearest ▪ FindClusters ▪ MeanShift ▪ FindShortestTour ▪ SequenceAlignment ▪ DistanceTransform ▪ GraphDistance ▪ CentralFeature