Wei Yao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Wei Yao has a total of 36 co-authors. Here are the top co-authors by number of shared publications:
- Chang-Jie Zhou
School of Sciences, Hebei University of Science and Technology, Shijiazhuang, China
4 shared publications - Zhichao Zhang
School of Mathematics and Statistics, Nanjing University of Information Science and Technology, Nanjing, China
4 shared publications - Guangxv Zhang
Nanjing Institute of Technology
3 shared publications Sang-Eon Han
3 shared publications- Manjun Cui
School of Mathematics and Statistics, Nanjing University of Information Science and Technology, Nanjing, China
2 shared publications - Qingguo Li
School of Mathematics Hunan University 410082 Changsha Hunan China
2 shared publications
Latest Publications
- 28 August 2026
- Published by Elsevier in Topology and its Applications
- Vol. 395
Quasi-similarity spaces as a kind of generalized metric spaces
- 28 August 2026
- Published by Elsevier in Topology and its Applications
- Vol. 395
Quasi-similarity spaces as a kind of generalized metric spaces
- 11 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 12, 270-282
A Novel Unified Extended Matrix for Graph Signal Processing: Theory and Application
Graph signal processing has become an essential tool for analyzing data structured on irregular domains. While conventional graph shift operators (GSOs) are effective for certain tasks, they inherently lack flexibility in modeling dependencies between non-adjacent nodes, limiting their ability to represent complex graph structures. To address this limitation, this study proposes the unified extended matrix (UEM) framework, which integrates the extended-adjacency matrix and the unified graph representation matrix via parametric design, enabling flexible adaptation to different graph structures and revealing more graph signal information. Theoretical analysis of the UEM is conducted, demonstrating positive semi-definiteness and eigenvalue monotonicity under specific conditions. Then, we proposed a graph Fourier transform based on UEM (UEM-GFT) that adaptively tunes spectral properties to enhance signal processing performance. Experimental results on synthetic and real-world datasets demonstrate that the UEM-GFT outperforms existing GSO-based methods in anomaly detection, achieving superior performance across diverse network topologies.
- 11 February 2026
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal and Information Processing over Networks
- Vol. 12, 270-282
A Novel Unified Extended Matrix for Graph Signal Processing: Theory and Application
Graph signal processing has become an essential tool for analyzing data structured on irregular domains. While conventional graph shift operators (GSOs) are effective for certain tasks, they inherently lack flexibility in modeling dependencies between non-adjacent nodes, limiting their ability to represent complex graph structures. To address this limitation, this study proposes the unified extended matrix (UEM) framework, which integrates the extended-adjacency matrix and the unified graph representation matrix via parametric design, enabling flexible adaptation to different graph structures and revealing more graph signal information. Theoretical analysis of the UEM is conducted, demonstrating positive semi-definiteness and eigenvalue monotonicity under specific conditions. Then, we proposed a graph Fourier transform based on UEM (UEM-GFT) that adaptively tunes spectral properties to enhance signal processing performance. Experimental results on synthetic and real-world datasets demonstrate that the UEM-GFT outperforms existing GSO-based methods in anomaly detection, achieving superior performance across diverse network topologies.
- 16 January 2026
- Published by Springer Nature in Computational and Applied Mathematics
- Vol. 45 (5) , 210
Distance-based variable precision fuzzy rough set over semi-grouping functions and its application in digital image contour extraction
- 16 January 2026
- Published by Springer Nature in Computational and Applied Mathematics
- Vol. 45 (5) , 210
Distance-based variable precision fuzzy rough set over semi-grouping functions and its application in digital image contour extraction
- 5 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 73 (1053587X) , 5286-5302
Graph Chirp Signal and Graph Fractional Vertex–Frequency Energy Distribution
Graph signal processing (GSP) has emerged as a powerful framework for analyzing data defined on irregular domains. In recent years, many classical techniques in signal processing (SP) have been successfully extended to GSP. Among them, chirp signals play a crucial role in various SP applications. However, graph chirp signals, despite their potential importance, have not yet been formally defined. Here, we define graph chirp signals and establish a comprehensive theoretical framework for their analysis.We propose the graph fractional vertex–frequency energy distribution (GFED), which provides a powerful tool for processing and analyzing graph chirp signals.We introduce the general fractional graph distribution (GFGD), a generalized vertex–frequency distribution, along with the reduced-interference GFED that effectively suppress cross-term interference and enhance signal clarity. Furthermore, we develop a novel GFED domain filtering method for detecting graph chirp signals, enabling robust signal analysis in noisy environments. Moreover, this method demonstrates strong practical potential in real-world denoising tasks, achieving performance superior to several state-of-the-art approaches. Finally, we explore the potential applications of graph chirp signals, highlighting their usefulness in denoising and compression.
- 5 December 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Signal Processing
- Vol. 73 (1053587X) , 5286-5302
Graph Chirp Signal and Graph Fractional Vertex–Frequency Energy Distribution
Graph signal processing (GSP) has emerged as a powerful framework for analyzing data defined on irregular domains. In recent years, many classical techniques in signal processing (SP) have been successfully extended to GSP. Among them, chirp signals play a crucial role in various SP applications. However, graph chirp signals, despite their potential importance, have not yet been formally defined. Here, we define graph chirp signals and establish a comprehensive theoretical framework for their analysis.We propose the graph fractional vertex–frequency energy distribution (GFED), which provides a powerful tool for processing and analyzing graph chirp signals.We introduce the general fractional graph distribution (GFGD), a generalized vertex–frequency distribution, along with the reduced-interference GFED that effectively suppress cross-term interference and enhance signal clarity. Furthermore, we develop a novel GFED domain filtering method for detecting graph chirp signals, enabling robust signal analysis in noisy environments. Moreover, this method demonstrates strong practical potential in real-world denoising tasks, achieving performance superior to several state-of-the-art approaches. Finally, we explore the potential applications of graph chirp signals, highlighting their usefulness in denoising and compression.
- 17 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 34 (1) , 263-272
Clustering Based on Transitive Closure of Distance Functions
Partition-based clustering is widely adopted for its simplicity and efficiency. However, it fails to capture non-convex, manifold-based, or linearly inseparable clusters, which commonly appear in real-world data, such as biological sequences, social networks, and medical images. This limitation severely restricts its application in recognizing clusters of arbitrary shapes. To address this challenge, we propose TC-KMeans, a $k$ -means method integrated with transitive closure. Its core innovation lies in the incorporation of transitive closure into distance functions: instead of relying solely on direct pairwise distances, TC-KMeans establishes sample associations by propagating relevance through transitive closure. This design allows it to treat highly correlated samples as a cohesive whole, even when the direct distance between data points is large. These improvements overcome the convex structure limitation of traditional partition-based methods, enabling TC-KMeans to detect arbitrary -shaped clusters while enhancing both clustering accuracy and completeness. To validate the proposed algorithm, we used Purity, AMI, ARI, FMI, and computation time as evaluation metrics, and conducted tests comparing it with 11 comparative algorithms on 20 synthetic datasets and 10 real-world datasets of different dimensions. Experimental results show that on synthetic non-convex datasets, TC-K-Means achieves a value of 1.0in key metrics for most datasets and has a shorter computation time than other algorithms; on real-world datasets, among the 12 algorithms, its average ranking in all metrics ranks first, outperforming traditional partition-based methods and density-based methods. In conclusion, TC-KMeans provides a practical solution for clustering complex non-convex data and expands the application scope of partition-based clustering algorithms.
- 17 November 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Fuzzy Systems
- Vol. 34 (1) , 263-272
Clustering Based on Transitive Closure of Distance Functions
Partition-based clustering is widely adopted for its simplicity and efficiency. However, it fails to capture non-convex, manifold-based, or linearly inseparable clusters, which commonly appear in real-world data, such as biological sequences, social networks, and medical images. This limitation severely restricts its application in recognizing clusters of arbitrary shapes. To address this challenge, we propose TC-KMeans, a $k$ -means method integrated with transitive closure. Its core innovation lies in the incorporation of transitive closure into distance functions: instead of relying solely on direct pairwise distances, TC-KMeans establishes sample associations by propagating relevance through transitive closure. This design allows it to treat highly correlated samples as a cohesive whole, even when the direct distance between data points is large. These improvements overcome the convex structure limitation of traditional partition-based methods, enabling TC-KMeans to detect arbitrary -shaped clusters while enhancing both clustering accuracy and completeness. To validate the proposed algorithm, we used Purity, AMI, ARI, FMI, and computation time as evaluation metrics, and conducted tests comparing it with 11 comparative algorithms on 20 synthetic datasets and 10 real-world datasets of different dimensions. Experimental results show that on synthetic non-convex datasets, TC-K-Means achieves a value of 1.0in key metrics for most datasets and has a shorter computation time than other algorithms; on real-world datasets, among the 12 algorithms, its average ranking in all metrics ranks first, outperforming traditional partition-based methods and density-based methods. In conclusion, TC-KMeans provides a practical solution for clustering complex non-convex data and expands the application scope of partition-based clustering algorithms.
- Open Access
- 29 October 2025
- Published by Hacettepe University in Hacettepe Journal of Mathematics and Statistics
- Vol. 54 (5) , 1730-1736
Lifts of $L$-valued powerset mapping systems
Motivated by Zadeh's extension of mappings, this paper introduces a concept of lifts of powerset mapping systems. In this framework, $L$-subsets are lifts of ordinary subsets, $L$-Zadeh functions are lifts of ordinary mappings, $L$-fuzzy rough set operators are lifts of classical rough set operators, $L$-interior/closure operators of $L$-topology are lifts of interior/closure operators of level topologies. Consequently, lifts of powerset mapping systems can be considered as a useful fuzzy structures for future study.
- 29 October 2025
- Published by Hacettepe University in Hacettepe Journal of Mathematics and Statistics
- Vol. 54 (5) , 1730-1736
Open AccessLifts of $L$-valued powerset mapping systems
Motivated by Zadeh's extension of mappings, this paper introduces a concept of lifts of powerset mapping systems. In this framework, $L$-subsets are lifts of ordinary subsets, $L$-Zadeh functions are lifts of ordinary mappings, $L$-fuzzy rough set operators are lifts of classical rough set operators, $L$-interior/closure operators of $L$-topology are lifts of interior/closure operators of level topologies. Consequently, lifts of powerset mapping systems can be considered as a useful fuzzy structures for future study.
- Open Access
Graph Embedding in the Graph Fractional Fourier Transform Domain
Spectral graph embedding plays a critical role in graph representation learning by generating low-dimensional vector representations from graph spectral information. However, the embedding space of traditional spectral embedding methods often exhibit limited expressiveness, failing to exhaustively capture latent structural features across alternative transform domains. To address this issue, we use the graph fractional Fourier transform to extend the existing state-of-the-art generalized frequency filtering embedding (GEFFE) into fractional domains, giving birth to the generalized fractional filtering embedding (GEFRFE), which enhances embedding informativeness via the graph fractional domain. The GEFRFE leverages graph fractional domain filtering and a nonlinear composition of eigenvector components derived from a fractionalized graph Laplacian. To dynamically determine the fractional order, two parallel strategies are introduced: search-based optimization and a ResNet18-based adaptive learning. Extensive experiments on six benchmark datasets demonstrate that the GEFRFE captures richer structural features and significantly enhance classification performance. Notably, the proposed method retains computational complexity comparable to GEFFE approaches.
Open AccessGraph Embedding in the Graph Fractional Fourier Transform Domain
Spectral graph embedding plays a critical role in graph representation learning by generating low-dimensional vector representations from graph spectral information. However, the embedding space of traditional spectral embedding methods often exhibit limited expressiveness, failing to exhaustively capture latent structural features across alternative transform domains. To address this issue, we use the graph fractional Fourier transform to extend the existing state-of-the-art generalized frequency filtering embedding (GEFFE) into fractional domains, giving birth to the generalized fractional filtering embedding (GEFRFE), which enhances embedding informativeness via the graph fractional domain. The GEFRFE leverages graph fractional domain filtering and a nonlinear composition of eigenvector components derived from a fractionalized graph Laplacian. To dynamically determine the fractional order, two parallel strategies are introduced: search-based optimization and a ResNet18-based adaptive learning. Extensive experiments on six benchmark datasets demonstrate that the GEFRFE captures richer structural features and significantly enhance classification performance. Notably, the proposed method retains computational complexity comparable to GEFFE approaches.
- Open Access
Multiple-Parameter Graph Fractional Fourier Transform: Theory and Applications
The graph fractional Fourier transform (GFRFT) applies a single global fractional order to all graph frequencies, which restricts its adaptability to diverse signal characteristics across the spectral domain. To address this limitation, in this paper, we propose two types of multiple-parameter GFRFTs (MPGFRFTs) and establish their corresponding theoretical frameworks. We design a spectral compression strategy tailored for ultra-low compression ratios, effectively preserving essential information even under extreme dimensionality reduction. To enhance flexibility, we introduce a learnable order vector scheme that enables adaptive compression and denoising, demonstrating strong performance on both graph signals and images. We explore the application of MPGFRFTs to image encryption and decryption. Experimental results validate the versatility and superior performance of the proposed MPGFRFT framework across various graph signal processing tasks.
Open AccessMultiple-Parameter Graph Fractional Fourier Transform: Theory and Applications
The graph fractional Fourier transform (GFRFT) applies a single global fractional order to all graph frequencies, which restricts its adaptability to diverse signal characteristics across the spectral domain. To address this limitation, in this paper, we propose two types of multiple-parameter GFRFTs (MPGFRFTs) and establish their corresponding theoretical frameworks. We design a spectral compression strategy tailored for ultra-low compression ratios, effectively preserving essential information even under extreme dimensionality reduction. To enhance flexibility, we introduce a learnable order vector scheme that enables adaptive compression and denoising, demonstrating strong performance on both graph signals and images. We explore the application of MPGFRFTs to image encryption and decryption. Experimental results validate the versatility and superior performance of the proposed MPGFRFT framework across various graph signal processing tasks.
- 25 July 2025
- Published by Elsevier in Fuzzy Sets and Systems
- Vol. 519
Logical distance-based fuzzy rough set model and its application in feature selection
- 25 July 2025
- Published by Elsevier in Fuzzy Sets and Systems
- Vol. 519
Logical distance-based fuzzy rough set model and its application in feature selection
- Open Access
- 24 June 2025
- Published by Hacettepe University in Hacettepe Journal of Mathematics and Statistics
- Vol. 54 (3) , 998-1020
Metric-based fuzzy rough sets for brain tumor magnetic resonance imaging classification
Computer-aided diagnosis systems help physicians diagnose diseases accurately at an early stage by automating preprocessing, image enhancement, and feature extraction, thus increasing patient survival rates. In this paper, we introduce an algorithm that leverages metric-based fuzzy positive regions to address the degradation of feature quality in brain tumor magnetic resonance imaging caused by inappropriate image enhancement. Employing sliding window blocks, the algorithm performs overlapping segmentation of magnetic resonance images and evaluates the membership of these blocks to decision classes by metric-based fuzzy positive regions. Blocks with the highest fuzzy positive region values are selected for multiple enhancement rounds, forming a candidate set that is sequentially integrated back into the original image. Finally, the features of the locally enhanced images are analyzed using the fuzzy positive region to generate the optimal feature set. To validate the effectiveness of the proposed algorithm, the features extracted using this method are compared with those extracted directly from the original image, globally enhanced images, and locally enhanced images processed based on similar fuzzy positive regions. The experimental results demonstrate that the proposed algorithm significantly outperforms the other three methods in various evaluation metrics, including the confusion matrix, classification accuracy, and the kappa coefficient.
- 24 June 2025
- Published by Hacettepe University in Hacettepe Journal of Mathematics and Statistics
- Vol. 54 (3) , 998-1020
Open AccessMetric-based fuzzy rough sets for brain tumor magnetic resonance imaging classification
Computer-aided diagnosis systems help physicians diagnose diseases accurately at an early stage by automating preprocessing, image enhancement, and feature extraction, thus increasing patient survival rates. In this paper, we introduce an algorithm that leverages metric-based fuzzy positive regions to address the degradation of feature quality in brain tumor magnetic resonance imaging caused by inappropriate image enhancement. Employing sliding window blocks, the algorithm performs overlapping segmentation of magnetic resonance images and evaluates the membership of these blocks to decision classes by metric-based fuzzy positive regions. Blocks with the highest fuzzy positive region values are selected for multiple enhancement rounds, forming a candidate set that is sequentially integrated back into the original image. Finally, the features of the locally enhanced images are analyzed using the fuzzy positive region to generate the optimal feature set. To validate the effectiveness of the proposed algorithm, the features extracted using this method are compared with those extracted directly from the original image, globally enhanced images, and locally enhanced images processed based on similar fuzzy positive regions. The experimental results demonstrate that the proposed algorithm significantly outperforms the other three methods in various evaluation metrics, including the confusion matrix, classification accuracy, and the kappa coefficient.