FREDPC: A Feasible Residual Error-Based Density Peak Clustering Algorithm With the Fragment Merging Strategy

Milan D Parmar, Wei Pang, Dehao Hao, Jianhua Jang, Wang Liupu, You Zhou (Corresponding Author)

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

The most common issues for many clustering algorithms include the slow convergence, requirement for pre-specification of a number of parameters, and the lack of robustness when dealing with anomalies. Recently, the density peak clustering (DPC) algorithm was proposed to discover the centers of clusters by finding the density peaks in a dataset based on their local densities. The DPC needs neither an iterative process nor a large number of parameters, and it supports a heuristic approach, known as the decision graph, to manually select cluster centroids. However, the selection of the key parameters of the DPC was not systematically investigated. In this paper, we propose the feasible residual error-based density peak clustering algorithm with the fragment merging strategy, where the local density within the neighborhood region is measured through the residual error computation and the resulting residual errors are then used to generate residual fragments for cluster formation. The model parameters are then able to be calculated from the equations with statistical theoretical justification. We also develop a semi-automatic cluster identification method to eliminate the iterative process of manual centroid selection. The robustness and effectiveness of the proposed algorithm compared to the DPC and other clustering algorithms are demonstrated through experiments on standard benchmark datasets. The proposed method named feasible residual error-based density peak clustering (FREDPC) algorithm with the fragment merging strategy only needs to perform in one single step without any iteration and thus it is fast and has a great potential to be applied on a wide range of applications.
Original languageEnglish
Pages (from-to)89789-89804
Number of pages16
JournalIEEE Access
Volume7
DOIs
Publication statusPublished - 3 Jul 2019

Fingerprint

Merging
Clustering algorithms
Specifications
Experiments

Keywords

  • clustering
  • density peak clustering
  • anomaly detection
  • residual error
  • residual fragment
  • Clustering

ASJC Scopus subject areas

  • Engineering(all)
  • Materials Science(all)
  • Computer Science(all)

Cite this

FREDPC: A Feasible Residual Error-Based Density Peak Clustering Algorithm With the Fragment Merging Strategy. / Parmar, Milan D; Pang, Wei; Hao, Dehao; Jang, Jianhua; Liupu, Wang; Zhou, You (Corresponding Author).

In: IEEE Access, Vol. 7, 03.07.2019, p. 89789-89804.

Research output: Contribution to journalArticle

Parmar, Milan D ; Pang, Wei ; Hao, Dehao ; Jang, Jianhua ; Liupu, Wang ; Zhou, You. / FREDPC: A Feasible Residual Error-Based Density Peak Clustering Algorithm With the Fragment Merging Strategy. In: IEEE Access. 2019 ; Vol. 7. pp. 89789-89804.
@article{10b831ff3cda42969c283860b2a56698,
title = "FREDPC: A Feasible Residual Error-Based Density Peak Clustering Algorithm With the Fragment Merging Strategy",
abstract = "The most common issues for many clustering algorithms include the slow convergence, requirement for pre-specification of a number of parameters, and the lack of robustness when dealing with anomalies. Recently, the density peak clustering (DPC) algorithm was proposed to discover the centers of clusters by finding the density peaks in a dataset based on their local densities. The DPC needs neither an iterative process nor a large number of parameters, and it supports a heuristic approach, known as the decision graph, to manually select cluster centroids. However, the selection of the key parameters of the DPC was not systematically investigated. In this paper, we propose the feasible residual error-based density peak clustering algorithm with the fragment merging strategy, where the local density within the neighborhood region is measured through the residual error computation and the resulting residual errors are then used to generate residual fragments for cluster formation. The model parameters are then able to be calculated from the equations with statistical theoretical justification. We also develop a semi-automatic cluster identification method to eliminate the iterative process of manual centroid selection. The robustness and effectiveness of the proposed algorithm compared to the DPC and other clustering algorithms are demonstrated through experiments on standard benchmark datasets. The proposed method named feasible residual error-based density peak clustering (FREDPC) algorithm with the fragment merging strategy only needs to perform in one single step without any iteration and thus it is fast and has a great potential to be applied on a wide range of applications.",
keywords = "clustering, density peak clustering, anomaly detection, residual error, residual fragment, Clustering",
author = "Parmar, {Milan D} and Wei Pang and Dehao Hao and Jianhua Jang and Wang Liupu and You Zhou",
note = "Funding Agency: 10.13039/501100001809-National Natural Science Foundation of China; Science and Technology Development Foundation of Jilin Province; Science Foundation of Education Department of Guangdong Province; Social Science Foundation of Education Department of Jilin Province; Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing, Xi’an University of Technology;",
year = "2019",
month = "7",
day = "3",
doi = "10.1109/ACCESS.2019.2926579",
language = "English",
volume = "7",
pages = "89789--89804",
journal = "IEEE Access",
issn = "2169-3536",
publisher = "IEEE Explore",

}

TY - JOUR

T1 - FREDPC: A Feasible Residual Error-Based Density Peak Clustering Algorithm With the Fragment Merging Strategy

AU - Parmar, Milan D

AU - Pang, Wei

AU - Hao, Dehao

AU - Jang, Jianhua

AU - Liupu, Wang

AU - Zhou, You

N1 - Funding Agency: 10.13039/501100001809-National Natural Science Foundation of China; Science and Technology Development Foundation of Jilin Province; Science Foundation of Education Department of Guangdong Province; Social Science Foundation of Education Department of Jilin Province; Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing, Xi’an University of Technology;

PY - 2019/7/3

Y1 - 2019/7/3

N2 - The most common issues for many clustering algorithms include the slow convergence, requirement for pre-specification of a number of parameters, and the lack of robustness when dealing with anomalies. Recently, the density peak clustering (DPC) algorithm was proposed to discover the centers of clusters by finding the density peaks in a dataset based on their local densities. The DPC needs neither an iterative process nor a large number of parameters, and it supports a heuristic approach, known as the decision graph, to manually select cluster centroids. However, the selection of the key parameters of the DPC was not systematically investigated. In this paper, we propose the feasible residual error-based density peak clustering algorithm with the fragment merging strategy, where the local density within the neighborhood region is measured through the residual error computation and the resulting residual errors are then used to generate residual fragments for cluster formation. The model parameters are then able to be calculated from the equations with statistical theoretical justification. We also develop a semi-automatic cluster identification method to eliminate the iterative process of manual centroid selection. The robustness and effectiveness of the proposed algorithm compared to the DPC and other clustering algorithms are demonstrated through experiments on standard benchmark datasets. The proposed method named feasible residual error-based density peak clustering (FREDPC) algorithm with the fragment merging strategy only needs to perform in one single step without any iteration and thus it is fast and has a great potential to be applied on a wide range of applications.

AB - The most common issues for many clustering algorithms include the slow convergence, requirement for pre-specification of a number of parameters, and the lack of robustness when dealing with anomalies. Recently, the density peak clustering (DPC) algorithm was proposed to discover the centers of clusters by finding the density peaks in a dataset based on their local densities. The DPC needs neither an iterative process nor a large number of parameters, and it supports a heuristic approach, known as the decision graph, to manually select cluster centroids. However, the selection of the key parameters of the DPC was not systematically investigated. In this paper, we propose the feasible residual error-based density peak clustering algorithm with the fragment merging strategy, where the local density within the neighborhood region is measured through the residual error computation and the resulting residual errors are then used to generate residual fragments for cluster formation. The model parameters are then able to be calculated from the equations with statistical theoretical justification. We also develop a semi-automatic cluster identification method to eliminate the iterative process of manual centroid selection. The robustness and effectiveness of the proposed algorithm compared to the DPC and other clustering algorithms are demonstrated through experiments on standard benchmark datasets. The proposed method named feasible residual error-based density peak clustering (FREDPC) algorithm with the fragment merging strategy only needs to perform in one single step without any iteration and thus it is fast and has a great potential to be applied on a wide range of applications.

KW - clustering

KW - density peak clustering

KW - anomaly detection

KW - residual error

KW - residual fragment

KW - Clustering

UR - http://www.scopus.com/inward/record.url?scp=85069792955&partnerID=8YFLogxK

UR - http://www.mendeley.com/research/fredpc-feasible-residual-errorbased-density-peak-clustering-algorithm-fragment-merging-strategy

U2 - 10.1109/ACCESS.2019.2926579

DO - 10.1109/ACCESS.2019.2926579

M3 - Article

VL - 7

SP - 89789

EP - 89804

JO - IEEE Access

JF - IEEE Access

SN - 2169-3536

ER -