【摘要】 针对现有的基于图的多视图聚类算法没有考虑不同视图的权重和视图数据存在噪声的问题，提出一种基于图的自适应加权多视图聚类算法。通过自适应邻域学习从原始数据中构造多个关系图，引入视图权重调节参数，减少噪声的影响；通过自适应学习将各个关系图融合成统一关系图，通过秩约束优化使数据点自动划分成所需的簇，从而得到聚类结果。在多视图数据集上的实验结果表明了该算法的有效性。更多还原

【Abstract】 Aimed at the existing graph-based multi-view clustering algorithms without considering the weight of different views and their problem of noise in view data, a graph-based adaptive weighted multi-view clustering algorithm is proposed. Multiple relational graphs were constructed from the original data through adaptive neighborhood learning, and the view weight adjustment parameters were introduced to reduce the influence of noise. Each graph was integrated into a unified graph by adaptive learning, and the data points were automatically divided into clusters by rank constraint optimization, so as to obtain the clustering results. Experimental results on multi-view data sets show the effectiveness of the proposed algorithm.更多还原