DenMune: Density peak based clustering using mutual nearest neighbors

Abbas, Mohamed A.; El-Zoghabi, Adel A.; Shoukry, Amin

doi:10.1016/j.patcog.2020.107589

Cited by 57 publications

(17 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Specifically, we select six datasets Jain, Flame, Aggregation, R15, Compound, and Pathbased in this sensitivity test because they vary in sizes (from 240 to 788), classes (from 2 to 7), and different density levels (1 or 2). As shown in Table 2, when we take different values of num (namely 5,8,10,12,15) , VDPC always obtains the best performance when num = 10. Moreover, when we set num=10, datasets Jain, Flame, Aggregation, R15 are found having one density level, while the rest having two density levels.…”

Section: Resultsmentioning

confidence: 95%

“…As illustrated in Figure 6, the overall range of representatives is [0.0354, 14.6423], and the range of the only gap is [1.375, 8.392] in this case (see (8)). Obviously, this gap divides the representatives into two different intervals (i.e., density levels): lower density level [0.0354, 1.375] and higher density level [8.392, 14.6423].…”

Section: Dividing Representatives Into Different Density Levelsmentioning

confidence: 85%

“…The density of the connected points shown in Figure 1 is different from the natural clusters on both ends, however, DBSCAN with fixed global parameter values may wrongly assign these connected points and consider all the data points in Figure 1 as one big cluster. Density peak clustering (DPC) is a recently proposed clustering algorithm [5], which receives more and more attention [6,7,8]. DPC makes a relatively novel assumption on the cluster formation that cluster centers are often surrounded by data points with lower density and they are also far away from other data points with higher density [5].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

VDPC: Variational Density Peak Clustering Algorithm

Wang¹,

Wang²,

Zhou³

et al. 2022

Preprint

View full text Add to dashboard Cite

The widely applied density peak clustering (DPC) algorithm makes an intuitive cluster formation assumption that cluster centers are often surrounded by data points with lower local density and far away from other data points with higher local density. However, this assumption suffers from one limitation that it is often problematic when identifying clusters with lower density because they might be easily merged into other clusters with higher density. As a result, DPC may not be able to identify clusters with variational density. To address this issue, we propose a variational density peak clustering (VDPC) algorithm, which is designed to systematically and autonomously perform the clustering task on datasets with various types of density distributions. Specifically, we first propose a novel method to identify the representatives among all data points and construct initial clusters based on the identified representatives for further analysis of the clusters' property. Furthermore, we divide all data points into different levels according to their local density and propose a unified clustering framework by combining the advantages of both DPC and DBSCAN. Thus, all the identified initial clusters spreading across different density levels are systematically processed to form the final clusters. To evaluate the effectiveness of the proposed VDPC algorithm, we conduct extensive experiments using 20 datasets including eight synthetic, six real-world datasets and six image datasets. The experimental results show that VDPC outperforms two classical algorithms (i.e., DPC and DBSCAN) and four state-of-the-art extended DPC algorithms.

show abstract

Section: Resultsmentioning

confidence: 95%

Section: Dividing Representatives Into Different Density Levelsmentioning

confidence: 85%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

VDPC: Variational Density Peak Clustering Algorithm

Wang¹,

Wang²,

Zhou³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…And this will require the development of models and methods for the synthesis of methods for the reception of visual spectrum data, obtained in real time [5,7,10,16,17,30]. Thus, the actual problem of the development of artificial intelligence is the development of the principles of computer perception of the outside world through the understanding of video data [1,2,4,9,11,22,24].…”

Section: Literature Reviewmentioning

confidence: 99%

Real-Time Sensing, Reasoning and Adaptation for Computer Vision Systems

Hrytsyk

Nazarkevych

2021

Lecture Notes in Computational Intelligence and Decision Making

View full text Add to dashboard Cite

Automatically recognition and classification of biological objects under microscope methods are shown in paper. Problem of separated of white-black and color images is studied. Method of separation of different type of objects (visual diapason of specter) with compare results is shown in the paper. Quality of segmentation methods analyses is presented in the paper. Schemes and table results of segmentation are exist. Methods of pattern recognition applicability for Computer Vision Systems of analysis and pattern recognition scenes in the visual spectrum are studied in the paper. The methods and algorithms can be used in Realtime Sensing, white-black and color patterns, reasoning and adaptation for Computer Vision Systems too. Example of such systems is the glasses for people with visual impairments; when the camera mounted in glasses receives and transmits environment data, and the contact plate with electrical leads via e-pulse transmits data to the eye retina. Author analyzed several pattern recognition methods that will allow to process data of the environment for the brain. This will make the visually impaired persons with sub reality vision better orientation in environment. Theoretical basic, algorithms and their compared for apply is presented in paper.

show abstract

“…The density of the connected points shown in Figure 1 is different from the natural clusters on both ends, however, DBSCAN with fixed global parameter values may wrongly assign these connected points and consider all the data points in Figure 1 as one big cluster. Density peak clustering (DPC) is a recently proposed clustering algorithm [23], which receives more and more attention [28,15,2]. DPC makes a relatively novel assumption on the cluster formation that cluster centers are often surrounded by data points with lower density and they are also far away from other data points with higher density [23].…”

Section: Introductionmentioning

confidence: 99%