streamMOA (version 0.1-0)

DSC_DenStream: DenStream Data Stream Clusterer

Description

Class implements the DenStream cluster algorithm for data streams.

Usage

DSC_DenStream(epsilon, mu = 1, beta = 0.001, lambda = 0.001,
    initPoints = 100, offline = 2, processingSpeed=100)

Arguments

epsilon

defines the epsilon-neighborhood in which the density of each micro-cluster is calculated (i.e., the maximal radius).

mu

minimum weight for core-micro-clusters (w>=mu). Range: 0 to max(double).

beta

multiplier for mu to detect outlier micro-clusters (w

lambda

decay constant.

initPoints

number of points to use for initialization via DBSCAN.

offline

offline multiplier for epsilion (between 2 and 20).

processingSpeed

Number of incoming points per time unit (between 1 and 1000).

Value

• An object of class DSC_DenStream (subclass of DSC, DSC_MOA, DSC_Micro)

Details

Interface to the DenStream implementation in MOA. DenStream applies weighted DBSCAN for reclustering (see Examples section below).

References

Cao F, Ester M, Qian W, Zhou A (2006). Density-Based Clustering over an Evolving Data Stream with Noise. In Proceedings of the 2006 SIAM International Conference on Data Mining, pp 326-337. SIAM.

Bifet A, Holmes G, Pfahringer B, Kranen P, Kremer H, Jansen T, Seidl T (2010). MOA: Massive Online Analysis, a Framework for Stream Classification and Clustering. In Journal of Machine Learning Research (JMLR).

See Also

DSC, DSC_Micro, DSC_MOA

Examples

set.seed(0)
# 3 clusters with 5% noise
dsd <- DSD_Gaussians(k=3, noise=0.05)

dsc <- DSC_DenStream(epsilon=.05)
cluster(dsc, dsd, 500)
dsc

# plot micro-clusters
plot(dsc, dsd)

# show macro-clusters (using density reachability)
plot(dsc, dsd, type="both")

# reclustering. DenStream micro-clusters with k-means instead
km <- DSC_Kmeans(k=3, weighted=TRUE)
recluster(km, dsc)
plot(km, dsd, type="both")