Fuzzy clustering ensemble techniques have been proven to yield more accurate and robust clustering results, with the mainstream methods relying on the fuzzy co-association (FCA) matrix. However, the inherent issues of low-value density and uniform dispersion in the FCA matrix significantly affect the performance of fuzzy clustering ensembles, an aspect that has been overlooked. To address this issue, we propose a novel framework for fuzzy clustering ensemble based on fuzzy matrix self-enhancement (FMSE). Specifically, we initially employ singular value decomposition to extract the principal components of the FCA matrix, thereby alleviating its low-value density. Second, on the basis of the criterion of fuzzy entropy, we measure the fuzziness of samples, design a metric for the fuzzy representativeness of samples, and incorporate it into a fusion-weighted structure for the reconstruction of the FCA matrix, mitigating uniform dispersion. Subsequently, on the basis of the self-enhanced fuzzy matrix model, we utilize a prototype diffusion approach to identify core samples and gradually allocate remaining samples to obtain a consensus clustering solution. Extensive comparative experiments on benchmark datasets against state-of-the-art clustering ensemble methods demonstrate the effectiveness and superiority of the proposed approach.

Clustering ensemble often provides robust and stable results without accessing original features of data, and thus has been widely studied. The conventional clustering ensemble methods often take the full multiple base partitions as inputs and provide a consensus clustering result. However, in many real-world applications, full base partitions are hard to obtain because some data may be missing in some base partitions. To tackle this problem, in this paper, we propose a novel partial clustering ensemble method, which takes the partial multiple base partitions as inputs. In this method, we simultaneously fill the missing values in the base partitions and ensemble them by fully considering the consensus and diversity. Moreover, to address the unreliability issue in the partial data scenario, we seamlessly plug it into a self-paced learning framework. The extensive experiments on benchmark data sets demonstrate the effectiveness and efficiency of the proposed method when handling incomplete data.

Zero-shot image recognition attempts to simulate the zero-shot learning mechanism of humans and recognizes the images of novel classes. It is crucial to learn transferable knowledge from seen classes and generalize it to unseen classes for image recognition in zero-shot learning (ZSL). Most existing ZSL methods extract visual features with pretrained backbone networks and learn transferable knowledge with the extracted visual features. However, the backbone networks are not pretrained for a special task, and the extracted visual features usually contain some distractive information for the ZSL task, which causes some discriminative information to be ignored or weakened and degrades the quality of knowledge learned from seen classes. Moreover, since visual samples of unseen classes are not obtainable, domain shift is another challenging problem. In this paper, we propose visual feature enhancement to learn more discriminative visual features via a graph convolutional network (GCN) and an attention mechanism for improving the quality of the learned transferable knowledge. Different from previous works, we explore the correlations between different latent visual patterns of an image and introduce GCN to enhance visual features. On the other hand, we take advantage of different learning mechanisms of GCN and MLP and propose dual classifier learning for improving the generalization and inference capabilities of our model. In end-to-end model training, the module of visual feature enhancement and the module of dual classifier learning are beneficial to each other via joint optimization. Finally, we perform extensive experiments in the ZSL setting and GZSL setting. The extensive experimental results verify the effectiveness and superiority of our method.

Cloud-based Online Social Networks (OSNs) make it easier for geographically dispersed users to communicate with each other. These users not only demand to quickly access their own data but also hope to access their friends' data with low latency. In order to solve the problem, it is necessary to design a replica placement strategy to manage data on large-scale social networks and reduce the data storage costs while meeting the access latency requirement. In this paper, we propose a novel genetic algorithm-based data placement strategy to find an optimal number of replicas for each user's data and their optimal location. The method can reduce the inter-server traffic load across servers and ensure that users can access data in a tolerable time. Experiments with real Facebook dataset demonstrate that our data placement strategy can significantly reduce the cost of data storage and inter-server traffic.