DISC: Density-Based Incremental Clustering by Striding over Streaming Data

Kim, Bogyeong; Koo, Kyoseung; Kim, Juhun; Moon, Bongki

doi:10.1109/icde51399.2021.00077

Cited by 4 publications

(3 citation statements)

References 31 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In addition, it has been confirmed in [19] that EDMStream is faster than these algorithms. DISC [29] can deal with metric spaces, but we demonstrated that AMD-DPC scales much better than DISC.…”

Section: Related Workmentioning

confidence: 88%

“…To conclude, any exact solutions are clearly expensive, and we need to design an approximate solution. Actually, many applications allow approximate results to achieve a fast update time [14], [17], [19], [21], [25]- [29]. To improve the update efficiency, we have to approximate the local density of each p ∈ P active , because its exact computation incurs Ω(n) time.…”

Section: Problem Definitionmentioning

confidence: 99%

“…• Semi-static [20], [21]: A state-of-the-art exact algorithm introduced in Section III, • EDMStream [19]: A state-of-the-art approximation algorithm (the inner parameter of EDMStream was set by following the original paper), • AMD-DPC (Approximate, Metric, and Dynamic DPC): Our algorithm presented in Section V, and • DISC [29]: A state-of-the-art density-based (DBSCANbased) algorithm for dynamic data. Although, as shown in Fig.…”

Section: Algorithms Our Experiments Evaluatedmentioning

confidence: 99%

See 2 more Smart Citations

Scalable and Accurate Density-Peaks Clustering on Fully Dynamic Data

Amagata

2022

2022 IEEE International Conference on Big Data (Big Data)

View full text Add to dashboard Cite

Clustering is a primitive and important operator that analyzes a given dataset to discover its hidden patterns and features. Because datasets are usually updated dynamically (i.e., it accepts continuous insertions and arbitrary deletions), analyzing such dynamic data is also an important topic, and dynamic clustering effectively supports it, but is a challenging problem. In this paper, we consider the problem of densitypeaks clustering (DPC) on dynamic data. DPC is one of the density-based clustering algorithms and attracts attention for many applications, due to its effectiveness. We investigate the hardness of this problem theoretically to measure the efficiencies of dynamic DPC algorithms. We prove that any exact solutions are costly, and propose an approximation algorithm to enable faster updates. We conduct experiments on real datasets, and the results confirm that our algorithm is much faster and more accurate than state-of-the-art.

show abstract

Section: Related Workmentioning

confidence: 88%