KNN_AGG {DDoutlier}

Aggregated k-nearest neighbors distance over different k's

Description

Function to calculate aggregated distance to k-nearest neighbors over a range of k's, as an outlier score. Suggested by Angiulli, F., & Pizzuti, C. (2002)

Usage

KNN_AGG(dataset, k_min = 5, k_max = 10)

Arguments

dataset

The dataset for which observations have an aggregated k-nearest neighbors distance returned

k_min

The k parameter starting the k-range

k_max

The k parameter ending the k-range. Has to be smaller than the number of observations in dataset and greater than or equal to k_min

Details

KNN_AGG computes the aggregated distance to neighboring observations by aggregating the results from k_min-NN to k_max-NN, such that if k_min=1 and k_max=3, results from 1NN, 2NN and 3NN are aggregated. A kd-tree is used for kNN computation, using the kNN function() from the 'dbscan' package. The KNN_AGG function is useful for outlier detection in clustering and other multidimensional domains.

Value

A vector of aggregated distance for observations. The greater the distance, the greater outlierness

Author(s)

Jacob H. Madsen

References

Angiulli, F., & Pizzuti, C. (2002). Fast Outlier Detection in High Dimensional Spaces. In Int. Conf. on Knowledge Discovery and Data Mining (SIGKDD). Helsinki, Finland. pp. 15-26. DOI: 10.1007/3-540-45681-3_2

Examples

# Create dataset
X <- iris[,1:4]

# Find outliers by setting a range of k's
outlier_score <- KNN_AGG(dataset=X, k_min=10, k_max=15)

# Sort and find index for most outlying observations
names(outlier_score) <- 1:nrow(X)
sort(outlier_score, decreasing = TRUE)

# Inspect the distribution of outlier scores
hist(outlier_score)