Article
Computer Science, Artificial Intelligence
Yiyang Liu, Hua Dai, Jiawei Li, Yu Chen, Geng Yang, Jun Wang
Summary: This paper proposes a BP-Model-based convoy mining algorithm that optimizes mining by adopting the divide-and-conquer methodology. The performance of the algorithms is evaluated on real-world datasets, showing that they are more efficient than existing convoy mining algorithms.
EXPERT SYSTEMS WITH APPLICATIONS
(2023)
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)
Review
Computer Science, Information Systems
Wided Oueslati, Sonia Tahri, Hela Limam, Jalel Akaichi
Summary: This paper investigates the literature on moving objects, trajectory data, and trajectory data warehouse modeling, and compares classical and ontological existing patterns. Through this research, the strong and limited contributions of these patterns are demonstrated.
COMPUTER SCIENCE REVIEW
(2023)
Article
Chemistry, Analytical
Yuanqiang Zhang, Weifeng Li
Summary: Maritime traffic pattern recognition is crucial for intelligent transportation services and ship monitoring. By utilizing AIS data and dynamic methods, it is possible to clean, compress, partition, and cluster maritime traffic data, enabling the recognition of traffic patterns.
Article
Computer Science, Theory & Methods
Ke Li, Hongyu Wang, Ziwen Chen, Lisi Chen
Summary: This study focuses on detecting co-movement patterns from trajectories. It proposes a simplified definition and a fast detection algorithm to produce real-time and high-quality results.
FUTURE GENERATION COMPUTER SYSTEMS-THE INTERNATIONAL JOURNAL OF ESCIENCE
(2023)
Article
Chemistry, Multidisciplinary
Alberto Blazquez-Herranz, Juan-Ignacio Caballero-Garzon, Albert Zilverberg, Christian Wolff, Alejandro Rodriguez-Gonzalez, Ernestina Menasalvas
Summary: Mobile devices with sensors generate geo-spatial data for future applications, with trajectory clustering being crucial for analyzing common point events. The CROSS-CPP project aims to provide tools for data storage and analysis, with an adapted Quickbundles algorithm showing superior performance in trajectory clustering experiments using various distance measures.
APPLIED SCIENCES-BASEL
(2021)
Article
Engineering, Multidisciplinary
Yinyin Zhang, Yongjun Li, Wenli Ji
Summary: This paper proposes a trajectory-based user movement pattern similarity measurement method for user identification. By finding frequent user stay points to represent daily activities and assigning global and local popularity to stay points, the method accurately characterizes the contribution of stay points to similarity measurement. Through a trajectory-oriented embedding method, the user-specific and site-independent movement patterns are preserved. The experiments show a significant improvement in matching users compared to the representative works.
IEEE TRANSACTIONS ON NETWORK SCIENCE AND ENGINEERING
(2023)
Article
Chemistry, Analytical
Shiting Ding, Zhiheng Li, Kai Zhang, Feng Mao
Summary: This study selects representative sequential pattern mining algorithms and evaluates their performance on taxi trajectory data. The results demonstrate that contiguous constraint-based algorithms show good performance in terms of balanced RAM consumption and execution time.
Article
Operations Research & Management Science
Behnam Tavakkol, Myong K. Jeong, Susan L. Albin
Summary: This paper introduces two new clustering validity indices for uncertain data, which perform better in validating clusters of uncertain data objects and are robust to outliers. The proposed indices use probabilistic distance measures to capture the distance between uncertain objects and can handle instances where existing indices may fail to detect the correct number of clusters.
ANNALS OF OPERATIONS RESEARCH
(2021)
Article
Computer Science, Artificial Intelligence
Kexin Chu, Min Zhang, Yaling Xun, Jifu Zhang
Summary: In this paper, a new clustering approach for mixed attribute data is proposed, which effectively reduces the similarity difference and inclination of similarity measure superposition by using a hybrid similarity measure and a calculation formula of similarity mean. This approach avoids artificial setting of similarity threshold parameters.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2023)
Article
Computer Science, Information Systems
Krishna Kumar Sharma, Ayan Seal
Summary: Uncertain data clustering is a crucial task in machine learning and pattern recognition. Multi-view clustering has gained attention for producing good results compared to single-view clustering, with the introduction of a self-adaptive mixture similarity measure (SAM) to address limitations of existing similarity measures. Experimental results show that SAM outperforms state-of-the-art methods in grouping uncertain data.
INFORMATION SCIENCES
(2021)
Article
Computer Science, Information Systems
Md Mostafizer Rahman, Yutaka Watanobe, Taku Matsumoto, Rage Uday Kiran, Keita Nakamura
Summary: This research proposes an educational data mining framework to support programming learning using unsupervised algorithms. By collecting and preprocessing problem-solving data from an online judge system, and applying MK-means clustering algorithm and frequent pattern growth algorithm for data mining, this framework effectively extracts useful features, patterns, and rules, providing suggestions for programming learning.
Article
Engineering, Marine
Chunhua Tang, Meiyue Chen, Jiahuan Zhao, Tao Liu, Kang Liu, Huaran Yan, Yingjie Xiao
Summary: The proposed FOLFST method addresses the challenges of ship trajectory clustering by effectively identifying both overall and local features of trajectories, surpassing traditional algorithms in terms of noise recognition, computational efficiency, and other key metrics.
Article
Computer Science, Artificial Intelligence
Xiaolei Ma, Enze Huo, Haiyang Yu, Honghai Li
Summary: Truck platooning involves a series of trucks driving closely through communication technologies, which can lead to significant energy savings. This study proposes data mining approaches to analyze truck platooning patterns and finds that adjusting speeds can improve platooning efficiency and fuel consumption.
KNOWLEDGE-BASED SYSTEMS
(2021)
Article
Chemistry, Analytical
Xuejing Di, Dong June Lew, Kwang Woo Nam
Summary: In this paper, algorithms are proposed to discover homogeneous groups from geo-tagged videos with view directions. The density clustering algorithm is also extended to support fields-of-view (FoVs) in the geo-tagged videos, and an optimization model based on a two-level grid-based index is proposed. Experimental evaluation on real and synthetic datasets demonstrates the efficiency and effectiveness of the proposed homogeneous-pattern-discovery approach.