Article
Computer Science, Artificial Intelligence
Xianxue Yu, Guoxian Yu, Jun Wang, Carlotta Domeniconi
Summary: This paper proposes a co-clustering ensemble approach based on multiple relevance measures to achieve good consensus solutions. The method evaluates the quality of base co-clusters and measures feature-to-object relevance, combining feature-to-feature and object-to-object relevance information. Experimental results demonstrate that the approach outperforms other related co-clustering ensembles and has reduced runtime cost.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Valerie Poulin, Francois Theberge
Summary: This paper introduces a family of graph partition similarity measures that consider the topology of the graph, contrasting with set partition similarity measures. Graph-aware and set partition measures exhibit opposite behaviors regarding resolution issues and offer complementary information for comparing graph partitions.
IEEE TRANSACTIONS ON PATTERN ANALYSIS AND MACHINE INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Kai Cong, Jin Yang, Hongjun Wang, Li Tao
Summary: This paper proposes a novel Gaussian gravitational model for cluster ensemble (GGMCE) that extends the processing scope of gravitational models to discrete data. The model transforms each base cluster into an object with mass using Newton's second law of motion. Additionally, a Gaussian agency strategy is introduced to explore the final cluster assignment, replacing the need for hyperparameters for the number of neighbors. Experimental results on various real-world datasets demonstrate the superiority of the proposed model over existing models.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Automation & Control Systems
Dong Huang, Chang-Dong Wang, Hongxing Peng, Jianhuang Lai, Chee-Keong Kwoh
Summary: Ensemble clustering based on fast propagation of cluster-wise similarities via random walks addresses the issues of lack of information at higher levels of granularity and neglect of multiscale relationships in current ensemble clustering research. By constructing a cluster similarity graph and conducting random walks, a new cluster-wise similarity matrix is derived to achieve an enhanced co-association matrix. The proposed approach demonstrates effectiveness and efficiency through extensive experiments on various real-world datasets.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Surender Singh, Abdul Haseeb Ganie
Summary: In this article, new similarity measures for picture fuzzy sets (PFSs) are proposed to distinguish inconsistent PFSs, with applications in pattern recognition and decision-making. The superiority of the proposed PFS similarity measures over existing ones is established through structured linguistic variables.
EXPERT SYSTEMS WITH APPLICATIONS
(2021)
Article
Biology
Luke T. Slater, John A. Williams, Andreas Karwath, Hilary Fanning, Simon Ball, Paul N. Schofield, Robert Hoehndorf, Georgios Gkoutos
Summary: This article explores the generation of multiple semantic similarity scores for patients based on different facets of their phenotypic manifestation, demonstrating the potential of faceted semantic clustering for gaining a deeper and more nuanced understanding of text-derived patient phenotypes and identifying clinically relevant phenotype relationships.
COMPUTERS IN BIOLOGY AND MEDICINE
(2021)
Article
Computer Science, Interdisciplinary Applications
Najmeh Akbarpour, Ebrahim Akbari, Homayun Motameni
Summary: Clustering is an important unsupervised method for organizing data, and the similarity of different clustering methods needs to be evaluated. This study proposes a novel Set Matching Index (SMI) based on extended derivation of Jaccard, Dice, and Cosine measures. The performance of the proposed indices is compared with popular external indices and a new index on ten real-world and synthetic datasets.
JOURNAL OF COMPUTATIONAL SCIENCE
(2023)
Article
Computer Science, Artificial Intelligence
Eva Blanco-Mallo, Laura Moran-Fernandez, Beatriz Remeseiro, Veronica Bolon-Canedo
Summary: This paper provides an overview of the use of distance measures in machine learning and data mining tasks. It examines the factors to consider when choosing the most appropriate measure and analyzes seven commonly used measures, exploring their properties and applications. The paper also conducts experiments to study their relationships, performance, and their performance in the presence of noise, as well as the execution time required by each measure.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Mansoureh Naderipour, Mohammad Hossein Fazel Zarandi, Susan Bastani
Summary: Most existing validity index methods focus primarily on one aspect of the graph, either the topological structure or the heterogeneous properties of vertices, neglecting the other. The proposed validity index addresses overlapping communities based on structural and attribute similarities, aiming for each community to have a densely connected sub-graph with homogeneous attribute values.
EXPERT SYSTEMS WITH APPLICATIONS
(2022)
Article
Computer Science, Artificial Intelligence
Hamid Rezaei, Negin Daneshpour
Summary: This article focuses on the challenge of determining the degree of similarity measurement in mixed data clustering. It proposes a more efficient method by innovating in three important aspects of clustering. The method considers both distance and the number of similar features for assigning data objects to clusters, and it outperforms other algorithms in terms of accuracy on three datasets.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
T. M. Athira, Sunil Jacob John, Harish Garg
Summary: This paper introduces Pythagorean fuzzy soft set, a structure that can better handle vagueness. It proposes similarity measures for this structure and conducts a comparative study. Additionally, a clustering algorithm based on the proposed similarity measure is introduced.
Article
Computer Science, Artificial Intelligence
Brindaban Gohain, Rituparna Chutia, Palash Dutta
Summary: This paper addresses the issue of similarity measurement between Pythagorean fuzzy sets. Two new similarity measures are proposed to effectively capture the similarity between Pythagorean fuzzy sets and reflect the newly defined containment property. The complement of Pythagorean fuzzy sets is also redefined, and some related results on similarity measures are presented. The proposed similarity measures are tested for applicability to medical diagnosis and clustering problems through hypothetical case studies.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Biology
Pallabi Patowary, Dhruba K. Bhattacharyya, Pankaj Barah
Summary: The identification of modules in a gene interaction network is crucial for understanding the overall network structure. This study introduces a novel similarity measure called SNMRS and applies it to extract and validate modules. The SNMRS method shows superior performance in terms of cluster-validity indices and reveals biologically relevant patterns from gene data sets.
COMPUTERS IN BIOLOGY AND MEDICINE
(2022)
Article
Computer Science, Interdisciplinary Applications
Yueyang Zhao, Lei Cui
Summary: This study presents a method to construct a fusion matrix using text similarity measures in order to address the deficiency in semantic representations of medical texts and achieve the clustering of PubMed database retrieval results. Experimental results show that the fusion matrix-based clustering method outperforms others in grouping the text sets, and clustering the training set is not necessary for improving clustering performance. The extracted high-frequency words in the category descriptions effectively distinguish the meanings of the categories. Therefore, the fusion matrix design is effective for clustering academic texts.
Article
Computer Science, Artificial Intelligence
Brindaban Gohain, Rituparna Chutia, Palash Dutta, Surabhi Gogoi
Summary: This paper introduces two new tools for decision-making problems, namely similarity measures between intuitionistic fuzzy sets. By incorporating parameters such as the difference of membership degrees, the difference of nonmembership degrees, the hesitancy factor, and the difference in the minimum and maximum of cross-evaluation factors, the proposed methods outperform existing ones and overcome their limitations.
INTERNATIONAL JOURNAL OF INTELLIGENT SYSTEMS
(2022)
Article
Computer Science, Hardware & Architecture
Guanyue Li, Xiwen Wei, Si Wu, Zhiwen Yu, Sheng Qian, Hau-San Wong
Summary: This article proposes an Adversarial Adaptive Interpolation-based AutoEncoder (AdvAI-AE) to address the mismatch issue caused by linear interpolation in the latent space. By training an additional interpolation correction module and applying prior matching, the performance gains in downstream tasks on benchmark datasets are significantly improved.
Article
Computer Science, Information Systems
Sihao Lin, Wenhao Wu, Si Wu, Yong Xu, Hau-San Wong
Summary: This paper proposes a method called Un2Reliab for generating realistic pedestrian instances in a semi-supervised setting. Un2Reliab uses an encoder-decoder-like generative network and a discriminative network to jointly train, and applies regularization and consistency to ensure the similarity and consistency of the synthesized instances with the real instances. Experimental results show that Un2Reliab is able to synthesize high-fidelity pedestrian instances and improves the state-of-the-art results on multiple semi-supervised pedestrian detection benchmarks.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
Article
Computer Science, Artificial Intelligence
Wenhao Wu, Qianfen Jiao, Hau-San Wong, Gaozhe Li, Si Wu
Summary: Sufficient labeled training data may not be available for pedestrian detection in many real-world scenes. In this paper, a Scene-adaptive Pseudo Annotation (SaPA) approach is proposed to improve the generalization performance and pseudo annotation quality for training a more precise and scene-specific pedestrian detector, by exploiting both source data with sufficient supervision and unannotated target data with domain-specific information.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Zhiwen Yu, Fengxu Ye, Kaixiang Yang, Wenming Cao, C. L. Philip Chen, Lianglun Cheng, Jane You, Hau-San Wong
Summary: In this article, a novel graph construction method for semisupervised classification is proposed. The method optimizes the similarity matrix in both label space and an additional subspace to achieve a better and more robust result compared to the original data space. Additionally, a high-quality subspace is obtained by learning the projection matrix.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Cheng Liu, Wenming Cao, Si Wu, Wenjun Shen, Dazhi Jiang, Zhiwen Yu, Hau-San Wong
Summary: This article introduces an asymmetric graph-guided multitask learning method combined with self-paced learning for survival analysis. The experimental results demonstrate that this method can improve the performance of survival analysis and achieve higher prediction accuracy compared to the previous state-of-the-art methods.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2022)
Article
Automation & Control Systems
Jian Zhong, Xiangping Zeng, Wenming Cao, Si Wu, Cheng Liu, Zhiwen Yu, Hau-San Wong
Summary: This article proposes a semisupervised multiple choice learning approach to enhance the predictive performance of classification models by jointly training a network ensemble on partially labeled data.
IEEE TRANSACTIONS ON CYBERNETICS
(2022)
Article
Automation & Control Systems
Kaixiang Yang, Zhiwen Yu, C. L. Philip Chen, Wenming Cao, Hau-San Wong, Jane You, Guoqiang Han
Summary: The research team proposed a hybrid classifier ensemble framework that includes data transformation and an adaptive undersampling process, which effectively solves the classification problem of imbalanced data. They also designed a progressive ensemble framework to improve the performance of the classifier ensemble.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Wenming Cao, Zhongfan Zhang, Cheng Liu, Rui Li, Qianfen Jiao, Zhiwen Yu, Hau-San Wong
Summary: In this paper, an enhanced deep clustering network (EDCN) is proposed, which consists of a Feature Extractor, a Conditional Generator, a Discriminator, and a Siamese Network. The EDCN utilizes adversarial training to generate two types of data and original data, which are used to train the Feature Extractor for effective latent representation learning. A Siamese network is adopted to generate realistic data using pseudo-labels for enhancing the Feature Extractor. Extensive experiments show the effectiveness and superiority of the proposed EDCN.
PATTERN RECOGNITION
(2022)
Article
Engineering, Biomedical
Cheng Liu, Si Wu, Dazhi Jiang, Zhiwen Yu, Hau-San Wong
Summary: Advances in high throughput experimental methods have led to the availability of diverse omic datasets in clinical analysis applications. Different types of omic data contribute to the understanding of disease progression. By integrating multiple omics data, the proposed View-aware Collaborative Learning (VaCoL) method can jointly boost the performance of survival prediction and subgroup identification.
IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Cheng Liu, Si Wu, Rui Li, Dazhi Jiang, Hau-San Wong
Summary: This work proposes a new approach for incomplete multi-view clustering by addressing the issue of missing instances through similarity graph completion and self-supervised multi-view graph completion algorithm. The inferred graph is leveraged in representation learning by incorporating constrained feature learning. The experimental results demonstrate the effectiveness of the proposed method.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2023)
Article
Computer Science, Artificial Intelligence
Kaixiang Yang, Zhiwen Yu, C. L. Philip Chen, Wenming Cao, Jane You, Hau-San Wong
Summary: In this paper, a weighted broad learning system (BLS) model is proposed to address the imbalance problem and outliers and noises in imbalanced data. By introducing a weighting mechanism and density calculation, the adaptive weighted BLS (AWBLS) and incremental weighted ensemble BLS (IWEB) are proposed, which demonstrate better classification performance in experimental results.
IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING
(2022)
Article
Computer Science, Artificial Intelligence
Quanpeng Song, Guanyue Li, Si Wu, Wenjun Shen, Hau-San Wong
Summary: In this paper, a Discriminator Feature-based Progressive Inversion (DFPI) model is proposed for GAN-based image reconstruction and enhancement. By learning the mapping relationship between discriminator features and generator features, high-quality reconstruction and enhancement are achieved.
KNOWLEDGE-BASED SYSTEMS
(2023)
Article
Computer Science, Artificial Intelligence
Zhen Xu, Si Wu, Qianfen Jiao, Hau-San Wong
Summary: This paper introduces a GAN-based generative model for accurately extracting and transferring makeup styles from reference facial images to target faces. The proposed model utilizes target-aware makeup style encoding and verification, and improves the accuracy and fidelity of makeup transfer through encoding the difference map and learning style consistency.
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Cheng Liu, Rui Li, Si Wu, Hangjun Che, Dazhi Jiang, Zhiwen Yu, Hau-San Wong
Summary: In this work, the authors study the incomplete multiview clustering (IMVC) scenario, where some instances are missing in certain views. They propose a graph propagation-based approach to address the issue of missing entries in the partial graph, and to exploit both consistency and complementary information. The proposed method outperforms state-of-the-art methods in extensive experiments.
IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS
(2023)
Article
Computer Science, Information Systems
Tianyi Chen, Si Wu, Xuhui Yang, Yong Xu, Hau-San Wong
Summary: Learning effective generative models for natural image synthesis is essential to reduce the reliance of deep models on massive training data. This research introduces a Semantic Regularized class-conditional Generative Adversarial Network, named SReGAN, to address the issue of limited data and labels in Fine-Grained Image Synthesis (FGIS) tasks. By incorporating additional discriminators and classifiers, the generator is encouraged to model both marginal and class-conditional data distributions, and to discover the differences between object categories, ultimately generating high-fidelity images on various FGIS benchmarks.
IEEE TRANSACTIONS ON MULTIMEDIA
(2022)
