Unsupervised Adaptive Embedding for Dimensionality Reduction

Deep Label Propagation with Nuclear Norm Maximization for Visual Domain Adaptation

Domain adaptation aims to leverage abundant label information from a source domain to an unlabeled target domain with two different distributions. Existing methods usually rely on a classifier to generate high-quality pseudo-labels for the target domain, facilitating the learning of discriminative features. Label propagation (LP), as an effective classifier, propagates labels from the source domain to the target domain by designing a smooth function over a similarity graph, which represents structural relationships among data points in feature space. However, LP has not been thoroughly explored in deep neural network-based domain adaptation approaches. Additionally, the probability labels generated by LP are low-confident and LP is sensitive to class imbalance problem. To address these problems, we propose a novel approach for domain adaptation named deep label propagation with nuclear norm maximization (DLP-NNM). Specifically, we employ the constraint of nuclear norm maximization to enhance both label confidence and class diversity in LP and propose an efficient algorithm to solve the corresponding optimization problem. Subsequently, we utilize the proposed LP to guide the classifier layer in a deep discriminative adaptation network using the cross-entropy loss. As such, the network could produce more reliable predictions for the target domain, thereby facilitating more effective discriminative feature learning. Extensive experimental results on three cross-domain benchmark datasets demonstrate that the proposed DLP-NNM surpasses existing state-of-the-art domain adaptation approaches.

Typicality-Aware Adaptive Similarity Matrix for Unsupervised Learning

Graph-based clustering approaches, especially the family of spectral clustering, have been widely used in machine learning areas. The alternatives usually engage a similarity matrix that is constructed in advance or learned from a probabilistic perspective. However, unreasonable similarity matrix construction inevitably leads to performance degradation, and the sum-to-one probability constraints may make the approaches sensitive to noisy scenarios. To address these issues, the notion of typicality-aware adaptive similarity matrix learning is presented in this study. The typicality (possibility) rather than the probability of each sample being a neighbor of other samples is measured and adaptively learned. By introducing a robust balance term, the similarity between any pairs of samples is only related to the distance between them, yet it is not affected by other samples. Therefore, the impact caused by the noisy data or outliers can be alleviated, and meanwhile, the neighborhood structures can be well captured according to the joint distance between samples and their spectral embeddings. Moreover, the generated similarity matrix has block diagonal properties that are beneficial to correct clustering. Interestingly, the results optimized by the typicality-aware adaptive similarity matrix learning share the common essence with the Gaussian kernel function, and the latter can be directly derived from the former. Extensive experiments on synthetic and well-known benchmark datasets demonstrate the superiority of the proposed idea when comparing with some state-of-the-art methods.

Multiview subspace clustering via low‐rank correlation analysis

In order to explore multi‐view data, existing low‐rank‐based multi‐view subspace clustering methods seek a common low‐rank structure from different views. However, in real‐world scenarios, each view will often hold complex structures resulting from noise or outliers, causing unreliable and imprecise graphs, which the previous methods cannot effectively ameliorate. This study proposes a new method based on low‐rank correlation analysis to overcome these limitations. Firstly, the canonical correlation analysis strategy is introduced to jointly find the low‐rank structures in different views. In order to facilitate a robust solution, a dual regularisation term is further introduced to find such low‐rank structures that maximise the correlation in respective views much better. Thus, a unifying clustering structure is then integrated into the model to characterise the connections between different views adaptively. In this way, noise suppression is achieved more effectively. Furthermore, we avoid the uncertainty of spectral post‐processing of the unifying clustering structure by imposing a rank constraint on its Laplacian matrix to obtain the clustering results explicitly, further enhancing computation efficiency. Experimental results obtained from several clustering and classification experiments performed using 3Sources, Caltech101‐20, 100leaves, WebKB, and Hdigit datasets reveal the proposed method's superiority over compared state‐of‐the‐art methods in Accuracy, Normalised Mutual Information, and F‐score evaluation metrics.