HDG-Tree: A Structure for Clustering High-dimensional Data Streams

Ren, Jiadong; Li, Lining; Xia, Yan

doi:10.1109/iita.2009.370

Cited by 1 publication

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Therefore, a recursive region partition technique is proposed in Sun et al [35] to find clusters in different subspaces. The synopsis structure of Lu et al [34] is also improved by Wang et al [36], Zhang et al [37], and Ren et al [38] to provide quick access to cluster information. In contrast to other works that cluster in the axis-parallel subspace, Borutta et al [39] and Borutta et al [40] performed high-dimensional clustering in an arbitrarily oriented subspace.…”

Section: General Findingsmentioning

confidence: 99%

“…Subspace Partition Clustering Algorithm for High-Dimensional Data Streams (SPDStream) [37] extends SOStream [36] by classifying each unit as dense, potentially dense, or adjacent and keeping all non-empty units in a CD-Tree lattice. High-dimensional Dense Grid Tree for Clustering High-Dimensional Data Streams Algorithm (HGStream) [38] is also an extension of SOStream [36] that finds dense units based on a grid density threshold and maintains all non-empty units in an HDG-Tree lattice. Moreover, Density-based Data Streams Subspace Clustering Algorithm over Weighted Sliding Windows (SDSStream) [26] uses a preceding conventional algorithm in its initialization stage to generate potential and outlier micro-clusters in the form of an exponential histogram of cluster features.…”

Section: Main Findings 321 Subspace Clustering Process In High-dimens...mentioning

confidence: 99%

“…Synopsis structures for subspace clustering algorithms in high-dimensional data streams include the fading cluster structure (FCS) [22,23,31], the fading temporal cluster feature (FTCF) [24,25], the fading cluster structure with histogram (FCH) [26][27][28], the projected micro-cluster (PMC) [30], exponential moving average micro-cluster (EAMC) [31], shared nearest-neighbour projection microcluster (SNN-PMC) [33], grid structure (GS) [33,34], monitoring lattice (ML) [36], CD-Tree lattice (CDTL) [37], HDG-Tree lattice (HDGTL) [38], micro-cluster (CCMicro) [39] and Concept [40]. These synopsis structures can be divided into two main categories: micro-cluster and grid structure.…”

Section: Synopsis Structure and Cluster Maintenance Mechanism In High...mentioning

confidence: 99%

See 2 more Smart Citations

Subspace Clustering in High-Dimensional Data Streams: A Systematic Literature Review

Ghani¹,

Aziz²,

AbdulKadir³

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

Clustering high dimensional data is challenging as data dimensionality increases the distance between data points, resulting in sparse regions that degrade clustering performance. Subspace clustering is a common approach for processing high-dimensional data by finding relevant features for each cluster in the data space. Subspace clustering methods extend traditional clustering to account for the constraints imposed by data streams. Data streams are not only high-dimensional, but also unbounded and evolving. This necessitates the development of subspace clustering algorithms that can handle high dimensionality and adapt to the unique characteristics of data streams. Although many articles have contributed to the literature review on data stream clustering, there is currently no specific review on subspace clustering algorithms in high-dimensional data streams. Therefore, this article aims to systematically review the existing literature on subspace clustering of data streams in high-dimensional streaming environments. The review follows a systematic methodological approach and includes 18 articles for the final analysis. The analysis focused on two research questions related to the general clustering process and dealing with the unbounded and evolving characteristics of data streams. The main findings relate to six elements: clustering process, cluster search, subspace search, synopsis structure, cluster maintenance, and evaluation measures. Most algorithms use a two-phase clustering approach consisting of an initialization stage, a refinement stage, a cluster maintenance stage, and a final clustering stage. The density-based top-down subspace clustering approach is more widely used than the others because it is able to distinguish true clusters and outliers using projected microclusters. Most algorithms implicitly adapt to the evolving nature of the data stream by using a time fading function that is sensitive to outliers. Future work can focus on the clustering framework, parameter optimization, subspace search techniques, memory-efficient synopsis structures, explicit cluster change detection, and intrinsic performance metrics. This article can serve as a guide for researchers interested in high-dimensional subspace clustering methods for data streams.

show abstract

Section: General Findingsmentioning

confidence: 99%

Section: Main Findings 321 Subspace Clustering Process In High-dimens...mentioning

confidence: 99%

Section: Synopsis Structure and Cluster Maintenance Mechanism In High...mentioning

confidence: 99%

See 1 more Smart Citation

Subspace Clustering in High-Dimensional Data Streams: A Systematic Literature Review

Ghani¹,

Aziz²,

AbdulKadir³

2023

Computers, Materials &Amp; Continua

View full text Add to dashboard Cite

show abstract

HDG-Tree: A Structure for Clustering High-dimensional Data Streams

Cited by 1 publication

References 7 publications

Subspace Clustering in High-Dimensional Data Streams: A Systematic Literature Review

Subspace Clustering in High-Dimensional Data Streams: A Systematic Literature Review

Contact Info

Product

Resources

About