The rapid growth in feature dimension may introduce implicit associations between features and labels in multi-label datasets, making the relationships between features and labels increasingly complex. Moreover, existing methods often adopt low-dimensional linear decomposition to explore the associations between features and labels. However, linear decomposition struggles to capture complex nonlinear associations and may lead to misalignment between the feature space and the label space. To address these two critical challenges, we propose innovative solutions. First, we design a random walk graph that integrates feature-feature, label-label, and feature-label relationships to accurately capture nonlinear and implicit indirect associations, while optimizing the latent representations of associations between features and labels after low-rank decomposition. Second, we align the variable spaces by leveraging low-dimensional representation coefficients, while preserving the manifold structure between the original high-dimensional multi-label data and the low-dimensional representation space. Extensive experiments and ablation studies conducted on seven benchmark datasets and three representative datasets using various evaluation metrics demonstrate the superiority of the proposed method.

The "Curse of dimensionality" is prevalent across various data patterns, which increases the risk of model overfitting and leads to a decline in model classification performance. However, few studies have focused on this issue in Partial Multi-label Learning (PML), where each sample is associated with a set of candidate labels, at least one of which is correct. Existing PML methods addressing this problem are mainly based on the low-rank assumption. However, low-rank assumption is difficult to be satisfied in practical situations and may lead to loss of high-dimensional information. Furthermore, we find that existing methods have poor ability to identify positive labels, which is important in real-world scenarios. In this paper, a PML feature selection method is proposed considering two important characteristics of dataset: label relationship's noise-resistance and label connectivity. Our proposed method utilizes label relationship's noise-resistance to disambiguate labels. Then the learning process is designed through the reformed low-rank assumption. Finally, representative labels are found through label connectivity, and the weight matrix is reconstructed to select features with strong identification ability to these labels. The experimental results on benchmark datasets demonstrate the superiority of the proposed method.

In data-driven fault diagnosis, feature selection not only reduces model complexity but also plays a pivotal role in improving prediction accuracy. Existing studies typically employ binary labels to indicate whether a sample exhibits a specific fault, yet this approach fails to capture the severity of the fault. Moreover, features extracted from vibration signals often contain significant noise and redundancy, which adversely affects the training of models. To address these issues, we propose an alternating optimization strategy for feature space and label space. In the label enhancement stage, continuous-valued labels capturing subtle differences in label values through sample-label similarity. In the feature selection stage, a regularization constraint based on feature coefficient correlation is introduced to achieve progressive elimination of redundant and noisy features. Through iterative optimization, label enhancement and feature selection mutually reinforce each other, leading to improved diagnostic accuracy. Building upon this strategy, this article proposes a novel multilabel feature selection method, label-enhanced feature selection (LEFS). Experimental results on real-world datasets for bearing and gearbox fault diagnosis validate the effectiveness and advantage of the LEFS method.