Article
Computer Science, Information Systems
Yuxian Duan, Changyun Liu, Song Li, Xiangke Guo, Chunlin Yang
Summary: This article introduces an improved affinity propagation algorithm based on optimization of preference (APBOP) for automatic clustering on high-dimensional data. APBOP aims to address the challenges of feature extraction from high-dimensional data and the sensitivity of the clustering performance to preference. The proposed method utilizes dimensionality reduction and preference optimization techniques to improve the effectiveness of affinity propagation.
INFORMATION SCIENCES
(2023)
Article
Computer Science, Artificial Intelligence
Ling Ding, Shifei Ding, Yanru Wang, Lijuan Wang, Hongjie Jia
Summary: The study proposes a new manifold p-spectral clustering algorithm (M-pSC) using path-based affinity measure to handle manifold data, constructing more accurate affinity matrix and improving clustering quality and robustness.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Wenyu Hao, Shanmin Pang, Zhikai Chen
Summary: The paper proposes a dynamic strategy to construct a common local representation and designs a fusion term to maximize the common structure of local and global representations for mutual reinforcement. By integrating local and global representation learning in a unified framework and utilizing an alternative iteration based optimization procedure, the algorithm demonstrates superiority over state-of-the-art methods through extensive experiments on benchmark datasets.
Article
Computer Science, Artificial Intelligence
Xiaoling Yao, Rongguo Zhang, Jing Hu, Kai Chang, Xiaojun Liu, Jian Zhao
Summary: An algorithm combining intrinsic dimension and local tangent space (IDLTS) is proposed for manifold spectral clustering image segmentation to address the high computational complexity and lack of local similarity information in constructing similarity matrix. Through experimental results, it is shown that the IDLTS method achieves good performance in terms of segmentation accuracy and time consumption by improving efficiency and incorporating more local similarity information.
Article
Computer Science, Artificial Intelligence
Paola Favati, Ornella Menchi
Summary: In this paper, a method for arbitrarily shaped clustering of points in a linear space is proposed. It integrates partitioning obtained by spectral clustering with a merging technique to reduce the criticalities caused by parameter choice. The experimental results show that the proposed method outperforms spectral clustering, DBSCAN, and Chameleon 2 algorithms for non-convex problems on both artificial and real-world datasets.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
Article
Computer Science, Information Systems
Issam Dagher, Sandy Mikhael, Oubaida Al-Khalil
Summary: Clustering is an important technique in data mining that separates data points into different groups based on similarity. Gabor face clustering using affinity propagation and structural similarity index outperforms other well known clustering algorithms in experimental results.
MULTIMEDIA TOOLS AND APPLICATIONS
(2021)
Article
Computer Science, Hardware & Architecture
Guoqiu Wen, Yonghua Zhu, Linjun Chen, Mengmeng Zhan, Yangcai Xie
Summary: The proposed method considers both local and global structures for nonlinear clustering, achieving competitive clustering performance on real data sets compared to state-of-the-art methods.
Review
Computer Science, Artificial Intelligence
Jui-Hung Chang, Yin-Chung Leung
Summary: This paper proposes a novel framework for dynamic image clustering without prior knowledge of the cluster count. By using deep learning and coordinate learning models to project high dimensional data onto a two-dimensional space, a new clustering algorithm is introduced to evaluate and classify the projected data.
Article
Computer Science, Artificial Intelligence
Yanru Wang, Shifei Ding, Lijuan Wang, Ling Ding
Summary: An improved Density-based adaptive p-spectral clustering algorithm (DAPSC) is proposed to adjust the similarity between sample points by considering prior information and strengthen the local correlation between data points; combining the density canopy method to update the initial clustering center and the number of clusters weakens the algorithm sensitivity caused by the original p-spectral clustering.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2021)
Article
Computer Science, Artificial Intelligence
Zafaryab Rasool, Sunil Aryal, Mohamed Reda Bouadjenek, Richard Dazeley
Summary: Density Peak Clustering (DPC) is a popular clustering algorithm that uses pairwise similarity to detect arbitrary shaped clusters. However, it is not robust for datasets with different densities and is sensitive to scale changes in data representation. This paper proposes an effective data-dependent similarity measure called MP-Similarity, and integrates it into DPC to create MP-DPC. The experiments show that MP-DPC outperforms DPC with Euclidean distance and existing similarity measures, and is robust to changes in data scales.
PATTERN RECOGNITION
(2023)
Article
Computer Science, Artificial Intelligence
Juanjuan Luo, Huadong Ma, Dongqing Zhou
Summary: The paper proposes a spectral clustering method based on multiobjective evolutionary algorithm, effectively determining the nonzero entries and values in the similarity matrix through a phased approach. By optimizing diversity and similarity, it achieves a balance between time cost and clustering accuracy as demonstrated in experiments.
COMPLEX & INTELLIGENT SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Xiaoyun Chen, Qiaoping Wang, Shanshan Zhuang
Summary: The proposed ensemble dimension reduction method for subspace clustering based on spectral disturbance (SD-EPLSR) learns weight coefficients according to the theory of spectral disturbance, ensuring that clustering results on each data subset are close to the consensus clustering result and that data subsets with similar clustering results have approximate weights.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Environmental Sciences
Haimiao Ge, Liguo Wang, Haizhu Pan, Yuexia Zhu, Xiaoyu Zhao, Moqi Liu
Summary: This paper proposes an improved affinity propagation algorithm based on complex wavelet structural similarity index and local outlier factor for clustering hyperspectral images. Experimental results show that the proposed method can improve the performance of traditional affinity propagation and provide competitive clustering results.
Article
Computer Science, Hardware & Architecture
Qifen Yang, Ziyang Li, Gang Han, Wanyi Gao, Shuhua Zhu, Xiaotian Wu, Yuhui Deng
Summary: Spectral clustering algorithm has become popular in recent years for data clustering problems. However, its performance is affected by the quality of the similarity matrix and its stability is compromised by the use of the K-means algorithm. Therefore, FDAP-SC proposes a new approach that improves neighbor information retrieval and uses shared nearest neighbors for constructing the similarity matrix. It also calculates clustering centers through message passing between nodes. Experimental results show that FDAP-SC outperforms existing algorithms in terms of neighbor determination, handling complex shape datasets, and achieving high accuracy on various datasets.
JOURNAL OF SUPERCOMPUTING
(2022)
Article
Automation & Control Systems
Zongze Wu, Sihui Liu, Chris Ding, Zhigang Ren, Shengli Xie
Summary: This paper proposes a more efficient method to learn a graph similarity matrix, where points within the same class have larger similarities and points from different classes have smaller similarities. By utilizing self-representation coefficient matrix learning and optimizing the co-association matrix, the fast block-diagonal structure of the coefficient matrix is enhanced. The constructed affinity graphs can clearly reveal the intrinsic structures of the data sets, and experimental results show that the proposed method outperforms state-of-the-art methods.
IEEE TRANSACTIONS ON SYSTEMS MAN CYBERNETICS-SYSTEMS
(2021)
Article
Computer Science, Artificial Intelligence
Lingheng Meng, Shifei Ding, Yu Xue
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2017)
Review
Computer Science, Artificial Intelligence
Shifei Ding, Lingheng Meng, Youzhen Han, Yu Xue
COGNITIVE COMPUTATION
(2017)
Proceedings Paper
Automation & Control Systems
Lingheng Meng, Rob Gorbet, Dana Kulic
Summary: This paper proposes a new reinforcement learning algorithm LSTM-TD3, which addresses the issues of Partially Observable MDPs by introducing a memory component, showing significant advantages in experiments. This approach is capable of effectively handling missing and noisy observation data.
2021 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Lingheng Meng, Rob Gorbet, Dana Kulic
Summary: Multi-step methods in reinforcement learning have been shown to improve performance, but the specific contributions to performance enhancement are still not well understood. This study analyzes the effect of multi-step methods on alleviating the overestimation problem in deep reinforcement learning, proposing two improved methods, MDDPG and MMDDPG, based on Deep Deterministic Policy Gradient.
2020 25TH INTERNATIONAL CONFERENCE ON PATTERN RECOGNITION (ICPR)
(2021)
Article
Robotics
Lingheng Meng, Daiwei Lin, Adam Francey, Rob Gorbet, Philip Beesley, Dana Kulic
ACM TRANSACTIONS ON HUMAN-ROBOT INTERACTION
(2020)
Proceedings Paper
Computer Science, Artificial Intelligence
Kazumi Kumagai, Ikuo Mizuuchi, Meng Lingheng, Alexandru Blidaru, Philip Beesley, Dana Kulic
2018 27TH IEEE INTERNATIONAL SYMPOSIUM ON ROBOT AND HUMAN INTERACTIVE COMMUNICATION (IEEE RO-MAN 2018)
(2018)