Func.SNN: A function to calculate the shared nearest neighbors (SNN)

In HighDimOut: Outlier Detection Algorithms for High-Dimensional Data

Description

This function calculate the shared nearest neighbors (SNN). SNN is reported to be more robust than k nearest neighbors. Firstly, the k nearest neighbor distances for each observation is calculated. Then, the shared nearest neighbor similarity is calculated based on the result of k nearest neighbor. Note that k.nn should be greater than k.sel.

Usage

Func.SNN(data, k.nn, k.sel)

Arguments

data	is the data frame containing the observations (should be numeric data). Each row represents an observation and each variable is stored in one column.
k.nn	specifies the value used for calculating the shared nearest neighbors.
k.sel	specifies the number of shared nearest neighbors

Value

The function returns the matrix containing the indices of top k shared nearest neighbors for each observation

Examples

Func.SNN(data=TestData[,1:2], k.nn=5, k.sel=3)

Example output

    
X1    1  2  3
  1   2  8 18
  2   1  8 12
  3  20 25 15
  4   9  6 10
  5   7 16 56
  6   4  9 13
  7   5 16 15
  8  18  1  2
  9   4  6 10
  10  4  9 13
  11 14 23 24
  12  2 17  1
  13  6 10 22
  14 11 23 24
  15  3  7 16
  16  5  7 15
  17 12  1  2
  18  8  1  2
  19 21 15  1
  20  3 25  1
  21 19  3 15
  22  4 13 14
  23 11 14 24
  24 11 14 23
  25  3 20  8
  26 35 28 36
  27 39 51 30
  28 26 35 36
  29 44 52 48
  30 51 27 37
  31 41 32 33
  32 33 50 31
  33 32 50 31
  34 46 31 32
  35 26 28 36
  36 26 28 35
  37 45 27 30
  38 43 44 48
  39 27 49 51
  40 26 28 35
  41 31 47 32
  42 49 30 37
  43 38 44 48
  44 29 52 38
  45 37 27 30
  46 34 47 31
  47 41 46 31
  48 29 38 43
  49 39 42 27
  50 32 33 31
  51 27 30 39
  52 29 44 26
  53 11 14 23
  54 59  5  7
  55 58 11 53
  56  5 16  7
  57 53 58 11
  58 53 55 57
  59  5  7 16
  60 34 46 32

HighDimOut documentation built on May 2, 2019, 12:16 p.m.