Clustering by Fast Search and Find of Density Peaks with Data Field

Wang, Shuliang; Wang, Dakui; Li, Caoyuan; Li, Yan; Ding, Guofu

doi:10.1049/cje.2016.05.001

Cited by 89 publications

(58 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This article also compared the value of 3/ √ 2d 0 , about which in [23] it is argued that the threshold value d 0 should multiply with 3/ √ 2 based on the definition of Gaussian distribution and the 3-sigma rule [36].…”

Section: Natural City Aggregated By Maximum Entropy Methodsmentioning

confidence: 99%

“…However, such a study gives rise to two main challenges: one is whether the TIN is suitable to construct the basic geographic unit for defining the natural city; the other is whether the breakpoint is reasonable since the head/tail breaks method takes the mean value as the threshold to aggregate triangles as the boundary. For instance, Rodriguez and Laio used 20% as the distance threshold in a FSDP decision graph [22], Wang introduced a minimum entropy method integrated with data field theory to seek the threshold value at the nadir for clustering [23], which could be taken as a derivation from the principle of maximum entropy. This paper proposes a SA-MaxEnt method that integrates Reilly's Law of Retail Gravitation and the maximum entropy (MaxEnt) method to address these two problems ( Figure 2).…”

Section: Introductionmentioning

confidence: 99%

“…The maximum entropy principle indicates that the probability distribution which best represents the current state of knowledge is the one with the largest entropy [26], thus this article assumes that the entropy value series for the ratio of all the polygons' perimeters to all possible threshold values would achieve the maximum value at the optimal value. On the other hand, according to the 3σ principle of the Gaussian (normal) distribution presented by maximum entropy, the threshold range should be expanded to 3/ √ 2σ [23]. Admittedly, such a hypothesis is not capable of delimitating a strict urban boundary based on the exponential decay of services, which would need further exploration with a fuzzy measurement method or ancillary data.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Delimitating the Natural City with Points of Interests Based on Service Area and Maximum Entropy Method

Liu

Xia

et al. 2019

Entropy

View full text Add to dashboard Cite

The natural city, which is essential to understand urban physical scale and identify urban sprawling in urban studies, represents the urban functional boundaries of the city defined by human activities rather than the administrative boundaries. Most studies tend to utilize physical environment data such as street networks and remote sensing data to delimitate the natural city, however, such data may not match the real distribution of human activity density in the new cities or even ghost cities in China. This paper suggests aggregating the natural city boundary from the service area polygons of points of interest based on Reilly’s Law of Retail Gravitation and the maximum entropy method, since most points of interests provide services for surrounding communities, reflecting the vitality in a bottom-up way. The results indicate that the natural city defined by points of interests shows a high resolution and accuracy, providing a method to define the natural city with POIs.

show abstract

Section: Natural City Aggregated By Maximum Entropy Methodsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Delimitating the Natural City with Points of Interests Based on Service Area and Maximum Entropy Method

Liu

Xia

et al. 2019

Entropy

View full text Add to dashboard Cite

show abstract

“…The core idea of the clustering algorithm is to describe cluster centers on the features [23]. The thought of the algorithm about cluster centers has two characteristics: 1) The density of itself is very big, that it is surrounded by its neighbors whose density are not more than itself; 2) The "distance" between other data points with greater density are relatively large, in other words, distance of the two class centers is relatively far [24].…”

Section: Density Peak Clustering Algorithmmentioning

confidence: 99%

“…Different scale data sets use different sample density measurement criteria. When the sample scale is small, the cut-off distance d c which is selected subjectively has greater impact through sample density measure on clustering results [21]. After drawing a decision graph based on density and distance, it is needed to circle the relatively large density and distance of data points as an artificial cluster center [22].…”

Section: Introductionmentioning

confidence: 99%

Improved Density Peak Clustering Algorithm Based on Choosing Strategy Automatically for Cut-off Distance and Cluster Centre

Wang

Han

et al. 2018

Teh. vjesn.

View full text Add to dashboard Cite

Due to the defect of quick search density peak clustering algorithm required an artificial attempt to determine the cut-off distance and circle the clustering centres, density peak clustering algorithm based on choosing strategy automatically for cut-off distance and cluster center (CSA-DP) is proposed. The algorithm introduces the improved idea of determining cut-off distance and clustering centres, according to the approximate distance that maximum density sample point to minimum density sample point and the variation of similarity between the points which may be clustering centres. First, obtaining the sample point density according to the k-nearest neighbour samples and tapping the sample sorting of the distance to the maximum density point; then finding the turning position of density trends and determining the cutoff distance on the basis of the turning position; finally, in view of the density peak clustering algorithm, finding the data points which may be the centres of the cluster, comparing the similarity between them and determining the final clustering centres. The simulation results show that the improved algorithm proposed in this paper can automatically determine the cut-off distance, circle the centres, and make the clustering results become more accurate. In the end, this paper makes an empirical analysis on the stock of 147 bio pharmaceutical listed companies by using the improved algorithm, which provides a reliable basis for the classification and evaluation of listed companies. It has a wide range of applicability.

show abstract

Bottlenecks and Feasible Solutions of Data Field Clustering in Impact Factor, Time Resolution, Selecting Core Objects and Merging Process

Zhang

Wei

Shen

et al. 2019

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Clustering by Fast Search and Find of Density Peaks with Data Field

Cited by 89 publications

References 15 publications

Delimitating the Natural City with Points of Interests Based on Service Area and Maximum Entropy Method

Delimitating the Natural City with Points of Interests Based on Service Area and Maximum Entropy Method

Improved Density Peak Clustering Algorithm Based on Choosing Strategy Automatically for Cut-off Distance and Cluster Centre

Bottlenecks and Feasible Solutions of Data Field Clustering in Impact Factor, Time Resolution, Selecting Core Objects and Merging Process

Contact Info

Product

Resources

About