STATE: A Clustering Algorithm Focusing on Edges Instead of Centers

Abstract

With the expansion of data scale and the increase in data complexity, it is particularly important to accurately identify clusters and efficiently save clustering results. To address this, we propose a novel clustering algorithm, Shape clustering based on data field (STATE), which can quickly identify clusters of arbitrary shapes and greatly reduce the storage space of clustering results in any datasets without reducing the accuracy. STATE mainly focuses on finding the edges of clusters and directions of edges instead of clustering centers through the data field. The results of STATE are presented as the edges of clusters without data objects inside clusters and without noise. Extensive experiments show that STATE can recognize complex data distribution in noisy environments without discrimination and greatly save the storage space of clustering results. When it is applied in a real‐world scene, facial feature extraction, STATE can recognize eyes, nose, mouth, eyebrows and facial contours automatically without calibrating key features or training. Using the extracted facial features, we achieve facial recognition with high accuracy.

Adaptive Tensor Rank Approximation for Multi-View Subspace Clustering

“…In [21], Jiang et al solved the low-rank matrix completion problem via nuclear norm minimization and Frobenius norm minimization function. Based on the two typical spectral-type single-view subspace clustering models, sparse subspace clustering (SSC) [22] and low-rank representation model (LRR) [11], some multi-view subspace clustering models have been proposed [3], [10], [23]- [25]. Xia et al [23] proposed a Markov chain based multi-view spectral clustering (RMSC) through a sparse low-rank decomposition.…”

“…Cao et al [3] maximized the diverse information among multi-views and explored the complementarity among multi-view data using the proposed method called Hilbert Schmidt independence criterion (HSIC). Wang et al [25] adopted the exclusivity and consistency regularizers to capture the consistent and exclusive information. Wang et al [10] proposed to use the intact space learning method for MSC problem.…”

“…[3], [5], [25], t-SVD- [8], HLR- [12], -MSC [13], [2], and [30]. Among these eleven methods, the first eight methods are convex, and the last three methods are non-convex.…”

Reverse-Nearest-Neighbor-Based Clustering by Fast Search and Find of Density Peaks

“…Researchers have proposed a variety of clustering algorithms based on different ideas, which can be broadly classified as partition-based [6], [7], grid-based [8], hierarchy-based [9]- [11], density-based [12], [13], and graphbased [14]. Specifically, the algorithm of clustering by fast search and find of density peaks (CFSFDP) [15], proposed in Science in 2014, is a density-based clustering algorithm.…”