Baohong Liu
Research Subjects
Baohong Liu has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
Renxian Zeng
School of Information Engineering, Jiangxi University of Water Resources and Electric Power Nanchang China
2 shared publicationsJiazhen Hou
School of Information Engineering, Jiangxi University of Water Resources and Electric Power Nanchang China
1 shared publication- Li Lv
School of Information Engineering, Nanchang Institute of Technology, Nanchang, China
1 shared publication Lifeng Qian
School of Information Engineering Jiangxi University of Water Resources and Electric Power Nanchang China
1 shared publication- Longzhe Han
School of Information Engineering, Nanchang Institute of Technology, Nanchang 330099, China
1 shared publication - Tanghuai Fan
College of Computer Information and Engineering, Nanchang Institute of Technology, Nanchang 330044, China
1 shared publication
Latest Publications
- 15 July 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (14)
Density Peaks Clustering Algorithm Based on Minimum Spanning Tree and Skeleton Construction
The density peaks clustering (DPC) algorithm suffers from two limitations on density‐inhomogeneous datasets: incorrect cluster‐center selection and the “domino effect” caused by its single‐step assignment strategy. To address these issues, this paper proposes DPC‐MSTSC, which integrates MST‐based local density refinement with cluster‐skeleton construction. First, a local minimum spanning tree built on K‐nearest neighbors is pruned to eliminate cross‐cluster edges, producing density estimates that accurately adapt to heterogeneous densities and enable reliable center identification. Second, K‐fully mutual nearest‐neighbor (K‐FMNN) points are identified as cluster cores, and cluster skeletons are constructed through second‐order KNN expansion, which simultaneously determines the cluster centers and the number of clusters without manual input. Noncore points are then assigned to the closest skeleton via an attractiveness function that combines shared‐neighbor overlap, spatial proximity, and density similarity, effectively blocking error propagation. Experiments on synthetic and real‐world datasets demonstrate that DPC‐MSTSC accurately identifies cluster structures and achieves highly competitive performance.
- 15 July 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (14)
Density Peaks Clustering Algorithm Based on Minimum Spanning Tree and Skeleton Construction
The density peaks clustering (DPC) algorithm suffers from two limitations on density‐inhomogeneous datasets: incorrect cluster‐center selection and the “domino effect” caused by its single‐step assignment strategy. To address these issues, this paper proposes DPC‐MSTSC, which integrates MST‐based local density refinement with cluster‐skeleton construction. First, a local minimum spanning tree built on K‐nearest neighbors is pruned to eliminate cross‐cluster edges, producing density estimates that accurately adapt to heterogeneous densities and enable reliable center identification. Second, K‐fully mutual nearest‐neighbor (K‐FMNN) points are identified as cluster cores, and cluster skeletons are constructed through second‐order KNN expansion, which simultaneously determines the cluster centers and the number of clusters without manual input. Noncore points are then assigned to the closest skeleton via an attractiveness function that combines shared‐neighbor overlap, spatial proximity, and density similarity, effectively blocking error propagation. Experiments on synthetic and real‐world datasets demonstrate that DPC‐MSTSC accurately identifies cluster structures and achieves highly competitive performance.
- 24 March 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (7)
A Variable Density‐Based Clustering Algorithm Based on Reverse Nearest Neighbors
Traditional density‐based clustering algorithms often fail to identify cluster boundaries when dealing with variable density data, resulting in poor clustering outcomes. In this paper, a reverse nearest neighbor variable density clustering algorithm (RNNVDBC) with only a single undetermined parameter () is proposed to mitigate the issue. The proposed algorithm introduces a method based on reverse nearest neighbors (RNN) and density partition approach to accurately discover clusters of varying densities, shapes, and sizes as well as noise. To verify its effectiveness, experiments on various artificial and real‐world datasets with different densities are employed using RNNVDBC and the counterpart algorithms. The clustering validity metrics of adjusted rand index (ARI), normalized mutual information (NMI) and purity are implemented to evaluate their performance. The results demonstrate that the proposed algorithm is more effective to get the number and boundaries of clusters for varying density datasets better maintaining the integrity and separation between clusters.
- 24 March 2026
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 38 (7)
A Variable Density‐Based Clustering Algorithm Based on Reverse Nearest Neighbors
Traditional density‐based clustering algorithms often fail to identify cluster boundaries when dealing with variable density data, resulting in poor clustering outcomes. In this paper, a reverse nearest neighbor variable density clustering algorithm (RNNVDBC) with only a single undetermined parameter () is proposed to mitigate the issue. The proposed algorithm introduces a method based on reverse nearest neighbors (RNN) and density partition approach to accurately discover clusters of varying densities, shapes, and sizes as well as noise. To verify its effectiveness, experiments on various artificial and real‐world datasets with different densities are employed using RNNVDBC and the counterpart algorithms. The clustering validity metrics of adjusted rand index (ARI), normalized mutual information (NMI) and purity are implemented to evaluate their performance. The results demonstrate that the proposed algorithm is more effective to get the number and boundaries of clusters for varying density datasets better maintaining the integrity and separation between clusters.
- 25 September 2020
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 33 (5)
Density peaks clustering based on k‐nearest neighbors sharing
Summary: The density peaks clustering (DPC) algorithm is a density‐based clustering algorithm. Its density peak depends on the density‐distance model to determine it. The definition of local density for samples used in DPC algorithm only considers distance between samples, while the environments of samples are neglected. This leads to the result that DPC algorithm performs poorly on complex data sets with large difference in density, flow pattern or cross‐winding. In the meantime, the fault tolerance of allocation strategy for samples is relatively poor. Based on the findings, this article proposes a density peaks clustering based on k‐nearest neighbors sharing (DPC‐KNNS) algorithm, which uses the similarity between shared neighbors and natural neighbors to define the local density of samples and the allocation. Comparison between theoretical analysis and experiments on various synthetic and real data reveal that the algorithm proposed in this article can discover the cluster center of complex data sets with large difference in density, flow pattern or cross‐winding. It can also provide effective clustering.
- 25 September 2020
- Published by Wiley in Concurrency and Computation: Practice and Experience
- Vol. 33 (5)
Density peaks clustering based on k‐nearest neighbors sharing
Summary: The density peaks clustering (DPC) algorithm is a density‐based clustering algorithm. Its density peak depends on the density‐distance model to determine it. The definition of local density for samples used in DPC algorithm only considers distance between samples, while the environments of samples are neglected. This leads to the result that DPC algorithm performs poorly on complex data sets with large difference in density, flow pattern or cross‐winding. In the meantime, the fault tolerance of allocation strategy for samples is relatively poor. Based on the findings, this article proposes a density peaks clustering based on k‐nearest neighbors sharing (DPC‐KNNS) algorithm, which uses the similarity between shared neighbors and natural neighbors to define the local density of samples and the allocation. Comparison between theoretical analysis and experiments on various synthetic and real data reveal that the algorithm proposed in this article can discover the cluster center of complex data sets with large difference in density, flow pattern or cross‐winding. It can also provide effective clustering.