Clustering ensemble based on sample's stability

Li, Feijiang; Qian, Yuhua; Wang, Jieting; Dang, Chuangyin; Jing, Liping

doi:10.1016/j.artint.2018.12.007

Cited by 95 publications

(43 citation statements)

References 52 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Some three-way clustering approaches were developed in recent years [42]- [45]. Recently, by observing that the samples can change between clusters in different clustering results, Li et al [46] introduced the stability of a sample in a clustering ensemble problem and presented a clustering ensemble algorithm based on sample's stability. The stability of a sample reflects its tendency of changing its cluster in the different clustering results, which gives a new idea for threeway clustering.…”

Section: Fig 1 the Framework Of Clustering Ensemblementioning

confidence: 99%

Three-Way Clustering Method Based on Stability Theory

Wang

Yang

2021

IEEE Access

View full text Add to dashboard Cite

Two-way clustering algorithms use one single set to represent a cluster, which cannot intuitively show the fringe objects of a cluster. Three-way clustering uses the core region and the fringe region to describe a cluster, which divide the universe into three disjoint sets to capture the three types of relationships between a cluster and a sample, namely, belong-to fully, belong-to partially and not belongto fully. One of the main problems of three-way clustering is to construct the core and the fringe of each cluster. In this paper, we propose a three-way clustering algorithm by using the stability of each sample. In the proposed algorithm, we use a set of base clustering results as inputs to obtain the samples' stability by using the co-association matrix and determinacy function. With this stability, the universe is divided into the universe core and the universe fringe according to a threshold for sample's stability. The universe core is constituted by the samples with high stability and is divided into the core region of each cluster by using kmeans algorithm. Whereas the universe fringe is constituted by the samples with low stability and is assigned into the fringe region of each cluster. Therefore, a three-way explanation of the cluster is naturally formed. The experimental results on UCI data sets show that the proposed algorithm is effective in revealing cluster structures. INDEX TERMS three-way decision; three-way clustering; clustering ensemble; sample's stability

show abstract

Section: Fig 1 the Framework Of Clustering Ensemblementioning

confidence: 99%

Three-Way Clustering Method Based on Stability Theory

Wang

Yang

2021

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Since cluster quality is an important characteristic for obtaining the higher final clustering result, there is a need to assess the quality of a component cluster. Some criteria have been developed to estimate the clustering quality [20], [21]. Normalized mutual information (NMI), a metric based on information theory, is employed in the criteria selection [20].…”

Section: Introductionmentioning

confidence: 99%

“…Some criteria have been developed to estimate the clustering quality [20], [21]. Normalized mutual information (NMI), a metric based on information theory, is employed in the criteria selection [20]. The average Silhouette index is used to evaluate clustering quality; it is based on intra-cluster and inter-cluster distances [21].…”

Section: Introductionmentioning

confidence: 99%

Semi-Supervised Selective Affinity Propagation Ensemble Clustering With Active Constraints

Lei

2020

IEEE Access

View full text Add to dashboard Cite

The semi-supervised selective affinity propagation ensemble clustering with active constraints (SSAPEC) method combines affinity propagation (AP) clustering algorithm with ensemble clustering and incorporates prior knowledge (almost a pairwise constraint) to improve the clustering performance. In this paper, the evaluation criterion coverage ratio, which is based on the structure of clustering result, is employed to judge the homogeneity of the component clusters without the assistance of ground-truth during the component cluster selection. And to solve the problem of random selection of pairwise constraints, an active learning strategy is applied to find most informative constraints based on the ensemble clustering result; these pairwise constraints selected by active learning strategy lead to a similarity matrix adjustment. Then, the final result is obtained by running the Average-link algorithm on the adjusted similarity matrix. The experimental results on fourteen datasets demonstrate that the SSAPEC method proposed in this paper outperforms other state-of-the-art ensemble clustering schemes. INDEX TERMS Semi-supervised learning, ensemble clustering, ensemble clustering selection, active learning strategy.

show abstract

“…An algorithm resamples different samples into the same cluster [23]; another one uses graph segmentation algorithm to solve clustering ensemble task [24]; furthermore, a mixed model is proposed [25]; Dempster-Shafer evidence theory is also introduced to clustering ensemble area [26]. Apart from all these, some researchers classify the sample into a transition sample and core sample and then use the core sample as the cluster segmentation basis [27,28] (see Figure 2). Besides these two aspects, prior knowledge is also applied to improve the clustering ensemble result [29,30].…”

Section: Introductionmentioning

confidence: 99%

Clustering Ensemble Model Based on Self-Organizing Map Network

Hua

2020

Computational Intelligence and Neuroscience

View full text Add to dashboard Cite

This paper proposes a clustering ensemble method that introduces cascade structure into the self-organizing map (SOM) to solve the problem of the poor performance of a single clusterer. Cascaded SOM is an extension of classical SOM combined with the cascaded structure. The method combines the outputs of multiple SOM networks in a cascaded manner using them as an input to another SOM network. It also utilizes the characteristic of high-dimensional data insensitivity to changes in the values of a small number of dimensions to achieve the effect of ignoring part of the SOM network error output. Since the initial parameters of the SOM network and the sample training order are randomly generated, the model does not need to provide different training samples for each SOM network to generate a differentiated SOM clusterer. After testing on several classical datasets, the experimental results show that the model can effectively improve the accuracy of pattern recognition by 4%∼10%.

show abstract

Clustering ensemble based on sample's stability

Cited by 95 publications

References 52 publications

Three-Way Clustering Method Based on Stability Theory

Three-Way Clustering Method Based on Stability Theory

Semi-Supervised Selective Affinity Propagation Ensemble Clustering With Active Constraints

Clustering Ensemble Model Based on Self-Organizing Map Network

Contact Info

Product

Resources

About