Density-ratio based clustering for discovering clusters with varying densities

Zhu, Ye; Ting, Kai Ming; Carman, Mark James

doi:10.1016/j.patcog.2016.07.007

Cited by 118 publications

(78 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…More recently, methods based on local, data‐dependent measures or procedures such as density ratio (Zhu, Ting, & Carman, ), local contrast (Chen, Ting, Washio, & Zhu, ), and multidimensional scaling (Zhu, Ting, & Angelova, ) have also been attempted to tackle the problem of clusters with highly varying densities.…”

Section: Discovering Clusters Of Different Densitiesmentioning

confidence: 99%

“…The method determines bins into which the mnearest neighbor distances are classified, using a reversible jump Markov Chain Monte Carlo strategy; m-nearest neighbor distances that fall into the same bin are then considered similar. More recently, methods based on local, data-dependent measures or procedures such as density ratio (Zhu, Ting, & Carman, 2016), local contrast (Chen, Ting, Washio, & Zhu, 2018), and multidimensional scaling (Zhu, Ting, & Angelova, 2018) have also been attempted to tackle the problem of clusters with highly varying densities.…”

Section: Adapting Distance Measures To Different Density Levelsmentioning

confidence: 99%

See 1 more Smart Citation

Density‐based clustering

Campello

Kröger

Sander

et al. 2019

WIREs Data Min & Knowl

View full text Add to dashboard Cite

Clustering refers to the task of identifying groups or clusters in a data set. In density‐based clustering, a cluster is a set of data objects spread in the data space over a contiguous region of high density of objects. Density‐based clusters are separated from each other by contiguous regions of low density of objects. Data objects located in low‐density regions are typically considered noise or outliers. In this review article we discuss the statistical notion of density‐based clusters, classic algorithms for deriving a flat partitioning of density‐based clusters, methods for hierarchical density‐based clustering, and methods for semi‐supervised clustering. We conclude with some open challenges related to density‐based clustering. This article is categorized under: Technologies > Data Preprocessing Ensemble Methods > Structure Discovery Algorithmic Development > Hierarchies and Trees

show abstract

Section: Discovering Clusters Of Different Densitiesmentioning

confidence: 99%

Section: Adapting Distance Measures To Different Density Levelsmentioning

confidence: 99%

Density‐based clustering

Campello

Kröger

Sander

et al. 2019

WIREs Data Min & Knowl

View full text Add to dashboard Cite

show abstract

“…APSCAN utilizes Affinity Propagation technique to identify the cluster densities based on local data variations. A density based clustering technique aiding in the discovery of clusters with varying densities within a single dataset was proposed by Zhu et al in [12]. In-general, the input data is considered to contain data distributed in uniform densities.…”

Section: Related Workmentioning

confidence: 99%

“…Two approaches exist in literature to solve this issue, namely modifying the algorithm appropriately and rescaling the data. The technique proposed in [12] utilizes the latter by rescaling the data to appropriately identify the clusters. DBSCAN is applied to the rescaled data to identify clusters.…”

Section: Related Workmentioning

confidence: 99%

Scalable density based spatial clustering with integrated one-class SVM for noise reduction

Ahmed

Razak

2018

IJET

View full text Add to dashboard Cite

Information extraction from data is one of the key necessities for data analysis. Unsupervised nature of data leads to complex computational methods for analysis. This paper presents a density based spatial clustering technique integrated with one-class SVM, a machine learning technique for noise reduction, a modified variant of DBSCAN called NRDBSCAN. Analysis of DBSCAN exhibits its major requirement of accurate thresholds, absence of which yields suboptimal results. However, identifying accurate threshold settings is unattainable. Noise is one of the major side-effects of the threshold gap. The proposed work reduces noise by integrating a machine learning classifier into the operation structure of DBSCAN. Further, the proposed technique is parallelized using Spark architecture, thereby increasing its scalability and its ability to handle large amounts of data. Experiments and comparisons with similar techniques indicate high scalability levels and high homogeneity levels in the clustering process.

show abstract

“…However, most density-based methods have difficulties to detect all clusters when the clusters have large variations of densities (Ertöz et al 2003a;Zhu et al 2016). For example, DBSCAN (Ester et al 1996), which uses a global density threshold to discriminate cluster core points from noise, fails to identify all clusters in the presence of greatly varying densities (Zhu et al 2016).…”

Section: Introductionmentioning

confidence: 99%

Local contrast as an effective means to robust clustering against varying densities

et al. 2018

Self Cite

View full text Add to dashboard Cite

Most density-based clustering methods have difficulties detecting clusters of hugely different densities in a dataset. A recent density-based clustering CFSFDP appears to have mitigated the issue. However, through formalising the condition under which it fails, we reveal that CFSFDP still has the same issue. To address this issue, we propose a new measure called Local Contrast, as an alternative to density, to find cluster centers and detect clusters. We then apply Local Contrast to CFSFDP, and create a new clustering method called LC-CFSFDP which is robust in the presence of varying densities. Our empirical evaluation shows that LC-CFSFDP outperforms CFSFDP and three other state-of-the-art variants of CFSFDP.

show abstract

Density-ratio based clustering for discovering clusters with varying densities

Cited by 118 publications

References 16 publications

Density‐based clustering

Density‐based clustering

Scalable density based spatial clustering with integrated one-class SVM for noise reduction

Local contrast as an effective means to robust clustering against varying densities

Contact Info

Product

Resources

About