Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Yi Yang has a total of 21 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Sparse-Gradient-Regularized Multi-View Clustering with Joint ℓ2,log Sparsity and a Tensor γ* Rank Surrogate

    by Yi Yang, Baojie Pan, Ming Yang

    Multi-view clustering seeks a consensus partition from heterogeneous feature views while retaining view-specific information. We propose SGLog-γ∗-MSC, a nonconvex multi-view subspace clustering framework that combines graph total variation, an ℓ2,log penalty, and a tensor γ∗ spectral penalty. These terms promote locally consistent self-representations, model sample-wise corruption, and capture shared low-rank structure across views, respectively. We derive an alternating augmented-Lagrangian algorithm with candidate-selection rules that return global minimizers for both nonconvex proximal subproblems. We also state explicit conditions under which accumulation points satisfy the KKT system and establish conditional whole-sequence convergence through the Kurdyka–Łojasiewicz framework. Hyperparameters are selected by a label-free protocol specified before the audited rerun; ground-truth labels are never used during selection. Across 20 recorded k-means++ initializations of each fixed embedding, the method attains NMI 0.8206±0.0155 on Yale and 0.8962±0.0211 on Scene-15. Relative to nine literature-reported baselines, these are the highest reported NMI values on the two datasets. On UCI digits, BBCSport, and ORL, the method reaches NMI 0.9837, 0.9634, and 0.9892, respectively, ranking second only to HLR-M2VS in the descriptive cross-paper comparison. Component-wise ablation identifies the ℓ2,log term as the largest and most consistent contributor, while the effects of the γ∗ surrogate and sparse-gradient term depend on the dataset. Compared with the tensor nuclear norm, the γ∗ surrogate improves performance on BBCSport, ORL, and UCI digits; sensitivity analysis shows that an interior γ also improves performance on Scene-15. Together, these results support combining robust error modeling with local and shared structural regularization while emphasizing the dataset dependence of individual components.

  • Open Access

    Porcellio scaber Algorithm with t-Distributed Elite Mutation for Global Optimization

    by Chao Zhang, Yi Yang

    As the convergence accuracy of the Porcellio scaber algorithm (PSA) is low, this study proposes an improved algorithm based on the t-distribution elite mutation mechanism. First, the improved algorithm applies a t-distribution mutation to each dimension of the optimal solution of each generation. Using the dominant information of the optimal solution and t-distribution characteristics, the result of the mutation is then employed as the updated location of the selected porcellio; thus, the algorithm enhances the ability to jump out of local extreme values and improves the convergence speed. Second, the updated iterative rule of PSA may lead to the loss of information of the elite porcellio in the last generation. To solve this problem, the judgment mechanisms of the current and previous optimal solutions are included in the algorithm process. Finally, dynamic self-adaptive improvement is applied to the weight allocation parameters of PSA. Simulation results on 24 benchmark functions show that the improved algorithm has significant advantages in convergence accuracy, convergence speed, and stability when compared with basic PSA, PSO, GSA, and FPA. This indicates that the improved algorithm has certain advantages in terms of optimization. An optimal solution with good practicability is obtained by solving three practical engineering problems: three-bar truss, welded beam, and tension/compression-spring design.


  • A Novel Regularized Model for Third-Order Tensor Completion

    by Yi Yang, Lixin Han, Yuanzhen Liu, Jun Zhu, Hong Yan

    Inspired by the accuracy and efficiency of the $\gamma$ -norm of a matrix, which is closer to the original rank minimization problem than nuclear norm minimization (NNM), we generalize the $\gamma$ -norm of a matrix to tensors and propose a new tensor completion approach within the tensor singular value decomposition (t-svd) framework. An efficient algorithm, which combines the augmented Lagrange multiplier and closed-resolution of a cubic equation, was developed to solve the associated nonconvex tensor multi-rank minimization problem. Experimental results show that the proposed approach has an advantage over current state of the art algorithms in recovery accuracy.


  • A robust personalized location recommendation based on ensemble learning

    by Jun Zhu, Lixin Han, Zhinan Gou, Yi Yang, Xiaofeng Yuan, Jingxian Li, Shu Li


  • Open Access

    A Method for Constructing Supervised Topic Model Based on Term Frequency-Inverse Topic Frequency

    by Zhinan Gou, Zheng Huo, Yuanzhen Liu, Yi Yang

    Supervised topic modeling has been successfully applied in the fields of document classification and tag recommendation in recent years. However, most existing models neglect the fact that topic terms have the ability to distinguish topics. In this paper, we propose a term frequency-inverse topic frequency (TF-ITF) method for constructing a supervised topic model, in which the weight of each topic term indicates the ability to distinguish topics. We conduct a series of experiments with not only the symmetric Dirichlet prior parameters but also the asymmetric Dirichlet prior parameters. Experimental results demonstrate that the result of introducing TF-ITF into a supervised topic model outperforms several state-of-the-art supervised topic models.


  • Open Access

    Personalized Recommendation via Trust-Based Diffusion

    by Yuanzhen Liu, Lixin Han, Zhinan Gou, Yi Yang

    The diffusion-based algorithm is a promising member of the family of recommendation algorithms. It makes recommendations through the diffusion process on user-object bipartite graphs. However, a user's taste is often influenced by his/her trusted friends in social networks. In this paper, we propose a new trust-based diffusion on tripartite graphs, which integrates explicit trust relations and implicit trust relations into the diffusion process. Explicit trust relations are obtained from the social networks while implicit trust relations are inferred from implicit feedback. The experimental results indicate that our proposed method has a remarkable improvement in accuracy, and even only implicit trust relations employed, that is, diffusion on user-object bipartite graphs, the recommendation accuracy is still enhanced. We further present a general framework of applying implicit trust relations and explicit trust relations into basic network-based diffusion method, which is more general and flexible, and specific parameters can be selected to meet actual requirements for different datasets and real-world online platforms.


  • Open Access

    Clustering Mixed Data Based on Density Peaks and Stacked Denoising Autoencoders

    by Baobin Duan, Lixin Han, Zhinan Gou, Yi Yang, Shuangshuang Chen

    With the universal existence of mixed data with numerical and categorical attributes in real world, a variety of clustering algorithms have been developed to discover the potential information hidden in mixed data. Most existing clustering algorithms often compute the distances or similarities between data objects based on original data, which may cause the instability of clustering results because of noise. In this paper, a clustering framework is proposed to explore the grouping structure of the mixed data. First, the transformed categorical attributes by one-hot encoding technique and normalized numerical attributes are input to a stacked denoising autoencoders to learn the internal feature representations. Secondly, based on these feature representations, all the distances between data objects in feature space can be calculated and the local density and relative distance of each data object can be also computed. Thirdly, the density peaks clustering algorithm is improved and employed to allocate all the data objects into different clusters. Finally, experiments conducted on some UCI datasets have demonstrated that our proposed algorithm for clustering mixed data outperforms three baseline algorithms in terms of the clustering accuracy and the rand index.


  • Open Access

    Tagrec-CMTF: Coupled Matrix and Tensor Factorization for Tag Recommendation

    by Yi Yang, Lixin Han, Zhinan Gou, Baobin Duan, Jun Zhu, Hong Yan

    In order to address data sparsity, missing value, and over-fitting problems in a social tagging system, a coupled matrix and tensor factorization (CMTF) method named Tagrec-CMTF for tag recommendation is proposed in this paper. In the CMTF method, we decompose the tag-item-user tensor joint with tag graph and two auxiliary matrices by using the CMTF, optimize the learning parameters with an alternating direction method of multipliers algorithm, and recommend the tag according to the predicted tensor. Our algorithm infuses the homogeneous and heterogeneous information of the tag and provides good prediction performance. Experiment results show that Tagrec-CMTF outperforms existing methods that do not utilize the homogeneous and heterogeneous information of the tag simultaneously.