ROBIN: ROBIN (ROBust INitialization)

Description Usage Arguments Details Value Author(s) References See Also Examples

View source: R/ROBIN.R

Description

ROBIN searches for k initial cluster seeds for k-means-based clustering methods.

Usage

1
ROBIN(D, data, k, mp = 10, critRobin = 1.05)

Arguments

D

A distance matrix calculated on data.

data

A data matrix with n observations and p variables.

k

The number of cluster centers to find.

mp

The number of the nearest neighbors to find dense regions by LOF, the default is 10.

critRobin

The cutoff value for LOF to determine the observations in the dense regions, the default is 1.05.

Details

The centers are the observations located in the most dense region and far away from each other at the same time. In order to find the observations in the highly dense region, ROBIN uses LOF (Local Outlier Factor, Breunig et al (2000)), see more details lof.

Value

centers

A numeric vector of k initial cluster centers corresponding to the k indices of observations.

lof

A real vector of local outlier factor values.

Author(s)

Sarka Brodinova <sarka.brodinova@tuwien.ac.at>

References

Hasan AM, et al. Robust partitional clustering by outlier and density insensitive seeding. Pattern Recognition Letters, 30(11), 994-1002, 2009.

See Also

lof, wrk

Examples

1
2
3
4
5
6
7
8
# generate data
d <- SimData(size_grp=c(40,40,40),p_inf=50,
p_noise=750,p_out_noise=75)
dat <- scale(d$x)

res <- ROBIN(D=dist(dat),data=dat,k=3)
km <- kmeans(x=dat,centers=dat[res$centers,])
table(d$y,km$cluster)

brodsa/wrsk documentation built on April 7, 2020, 6:12 a.m.