Review
Computer Science, Information Systems
Mahmoud A. Mahdi, Khalid M. Hosny, Ibrahim Elhenawy
Summary: In the era of big data, traditional clustering algorithms face high computational costs, making it challenging to accurately process massive amounts of data in crucial moments. Despite the development of different algorithms to facilitate clustering processes, there are still many difficulties when dealing with large data volumes.
Article
Engineering, Electrical & Electronic
Iresha Pasquel Mohottige, Hassan Habibi Gharakheili, Tim Moors, Vijay Sivaraman
Summary: With the surge in enrollments in universities worldwide, campus estate managers are seeking continuous data on attendance patterns to optimize the usage of classroom space. This paper proposes machine learning-based models to infer classroom occupancy from WiFi sensing infrastructure, and develops methods to map access points (APs) to classrooms and evaluate different algorithms for estimating room occupancy. The results show promising accuracy in mapping APs to classrooms and comparable estimation for room occupancy.
IEEE SENSORS JOURNAL
(2022)
Article
Computer Science, Information Systems
Junhua Wong, Vincenzo Piuri, Fabio Scotti, Qingxue Zhang
Summary: The Internet of Medical Things (IoMT) is driving the development of emerging smart health applications by streaming big data for data-driven innovations. However, the power hungriness of long-term data transmission is a critical obstacle in IoMT big data. To address this challenge, we propose a novel framework called IoMT big-data Bayesian-backward deep-encoder learning (IBBD) that utilizes deep autoencoder (AE) configurations to sparsify data and determine optimal tradeoffs between information loss and power overhead.
IEEE INTERNET OF THINGS JOURNAL
(2023)
Article
Computer Science, Artificial Intelligence
Samayveer Singh, Aridaman Singh Nandan, Geeta Sikka, Aruna Malik, Pradeep Kumar Singh
Summary: Internet of Things (IoT)-enabled Wireless Sensor Networks is a promising research domain and industrial trend. This paper proposes a genetic algorithm integrated with efficient clustering for power grid applications in IoT-enabled networks. The OptiGeA protocol is developed for cluster heads election based on density, distance, energy, and heterogeneous node's capacity. The investigation analysis of OptiGeA shows superior performance compared to state-of-the-art protocols in terms of stability period, system's residual energy, network lifetime, throughput, and number of clusters per round execution.
Article
Computer Science, Information Systems
Samayveer Singh, Aridaman Singh Nandan, Aruna Malik, Rajeev Kumar, Lalit K. Awasthi, Neeraj Kumar
Summary: This article proposes an optimized genetic algorithm-based method for sustainable and secure green data collection/transmission in IoT-enabled WSN in healthcare. The method optimizes intracluster distance, node's energy utilization, and reduces hop count. The data is encrypted using stream cipher and a pseudo-randomly generated security key for secure transmission. The proposed movable sink and data collection/transmission strategies shorten communication distance and diminish the hotspot problem. The incorporation of dynamic sensing range minimizes energy consumption. Simulation results show the superiority of the proposed protocol over existing protocols on various performance metrics.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Information Systems
Magda M. Madbouly, Saad M. Darwish, Noha A. Bagi, Mohamed A. Osman
Summary: This paper introduces an improved hierarchical distributed k-medoid clustering method specifically designed for spatial query analysis in big data. By utilizing the fuzzy k-medoid method, the efficiency and accuracy of clustering are improved by overcoming outliers and data uncertainties. The method consists of two phases, creating local clusters and aggregating them into final clusters.
Article
Computer Science, Information Systems
Dongwei Li, Shuliang Wang, Nan Gao, Qiang He, Yun Yang
Summary: This research proposes a novel approach to achieve cost-effective big data clustering in the cloud by training a regression model with sampling data, allowing k-means and EM algorithms to stop automatically when desired accuracy is obtained. Experiment results show high cost-effectiveness with k-means needing only 47.71-71.14% of the computation cost for 99% accuracy and EM needing 16.69-32.04%, potentially saving up to $94,687.49 for the government in each use case.
IEEE TRANSACTIONS ON CLOUD COMPUTING
(2022)
Article
Engineering, Electrical & Electronic
Jianbo Zhang, Subin Zhao, Zhuangzhuang Ye
Summary: This article presents a Spark-based parallel computing approach for viewshed analysis, which effectively handles high-resolution DEM data and improves processing efficiency and scalability for the XDraw algorithm.
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
(2021)
Article
Computer Science, Information Systems
Gazi M. E. Rahman, Khan A. Wahid
Summary: This article proposes a real-time lightweight dynamic clustering algorithm (LDCA) for a wireless sensor network with limited processing resources. The algorithm is based on the received signal strength indicator and signal-to-noise ratio of a long-range (LoRa) interface and its residual energy, reducing the energy requirement by 33% through reducing concurrent clusters and hops.
IEEE INTERNET OF THINGS JOURNAL
(2022)
Article
Computer Science, Information Systems
Yubo Wang, Shelesh Krishna Saraswat, Iraj Elyasi Komari
Summary: Ensemble clustering, which combines the results of multiple clustering methods, is a challenging research direction in data mining. This study introduces a parallel hierarchical clustering approach using divide-and-conquer strategy to achieve faster and more efficient ensemble clustering. A cluster consensus selection approach is proposed, which selects a subset of primary clusters to participate in the final consensus based on sample-cluster and cluster-cluster similarity. The proposed scheme also incorporates an unsupervised feature selection approach to remove irrelevant features. Extensive evaluations on datasets show that the proposed scheme outperforms state-of-the-art clustering methods, improving average performance by 6% to 24%.
JOURNAL OF KING SAUD UNIVERSITY-COMPUTER AND INFORMATION SCIENCES
(2023)
Article
Computer Science, Information Systems
Mohammad Sultan Mahmud, Joshua Zhexue Huang, Rukhsana Ruby, Alladoumbaye Ngueilbaye, Kaishun Wu
Summary: This paper proposes a distributed computing framework to tackle the challenging task of clustering a big distributed dataset. The approach uses multiple random samples to compute an ensemble result as an estimation of the true result of the dataset. The framework proves to be efficient and scalable in clustering big datasets.
IEEE TRANSACTIONS ON BIG DATA
(2023)
Article
Computer Science, Information Systems
Xiao Xue, Shuai Huangfu, Lejun Zhang, Shufang Wang
Summary: The paper discusses the categorization of internet business into pure online and O2O business, highlighting the need to improve O2O service recommendation through continuous feedback learning between cyber and social layers. The proposed research framework enhances user experience and helps O2O services escape big-data traps, as demonstrated by computational experiments.
IEEE TRANSACTIONS ON BIG DATA
(2021)
Article
Computer Science, Artificial Intelligence
Zuohong Xu, Zhou Zhang, Shilian Wang, Alireza Jolfaei, Ali Kashif Bashir, Ye Yan, Shahid Mumtaz
Summary: This article explores the issue of decentralized secondary users performing multiple channel sensing and access in cognitive radio networks under an unknown environment. By estimating and learning from big-data samples collected from wireless channels, online algorithms are proposed to reduce learning loss. The theoretical analysis and simulations show that the regret of the algorithms is both logarithmic in finite time and asymptotically.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2021)
Article
Computer Science, Interdisciplinary Applications
Rasim M. Alguliyev, Ramiz M. Aliguliyev, Lyudmila Sukhostat
Summary: This article introduces a new parallel batch clustering algorithm based on the k-means algorithm, which reduces computation complexity by splitting the dataset into multiple partitions and proposes a method to determine the optimal batch size. Experimental results show the practical applicability of this method for handling Big Data.
COMPUTERS & INDUSTRIAL ENGINEERING
(2021)
Article
Computer Science, Artificial Intelligence
Preeti Jha, Aruna Tiwari, Neha Bharill, Milind Ratnaparkhe, Mukkamalla Mounika, Neha Nagendra
Summary: The study proposes a novel Kernelized Scalable Random Sampling with Iterative Optimization Fuzzy c-Means (KSRSIO-FCM) clustering algorithm for efficiently clustering non-linear separable data in a Big Data framework. Experimental results demonstrate that the KSRSIO-FCM algorithm achieves significant improvements in time/space complexity and evaluation metrics compared to other scalable clustering algorithms.
IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE
(2021)
Article
Computer Science, Artificial Intelligence
Xinyuan Zhou, Yangli-ao Geng, Haomin Yu, Qingyong Li, Liangtao Xu, Wen Yao, Dong Zheng, Yijun Zhang
Summary: Lightning disaster poses a significant threat to human lives and industrial facilities. Data-driven lightning forecasting has proven effective in reducing such disaster losses. However, current methods fail to capture the long-range spatiotemporal dependencies within data. To address this issue, we propose a dual-source lightning forecasting network called LightNet+, which utilizes bi-directional spatiotemporal transformation to model long-range connections and extract historical trend information for accurate predictions.
APPLIED INTELLIGENCE
(2022)
Article
Meteorology & Atmospheric Sciences
Wei Wang, Xingqin An, Qingyong Li, Yangli-ao Geng, Haomin Yu, Xinyuan Zhou
Summary: In this study, two deep learning models, DeepPM and APTR, were constructed and trained to improve the forecasting effectiveness of numerical air quality models. The results showed that these models significantly outperformed the WRF-Chem numerical model in both short-term and medium-term forecasts.
ATMOSPHERIC RESEARCH
(2022)
Article
Physics, Multidisciplinary
Min Zheng, Yangliao Geng, Qingyong Li
Summary: This paper proposes a novel fine-grained image retrieval method that enhances the discriminative ability among different fine-grained classes by learning a global-local aware feature representation and exploring the intrinsic relationship of different parts via frequent pattern mining to obtain representative local features. Experimental results demonstrate the improved performance of fine-grained image retrieval with the proposed method.
Article
Geochemistry & Geophysics
Yang Liu, Wen Wang, Qingyong Li, Min Min, Zhigang Yao
Summary: The article introduces a new deformable convolutional cloud detection network, named DCNet, to enhance the adaptability of the model to cloud variations, which outperforms several state-of-the-art methods in experiments.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Article
Physics, Multidisciplinary
Lei Jia, Jianzhu Wang, Tianyuan Wang, Xiaobao Li, Haomin Yu, Qingyong Li
Summary: This paper presents a large-scale vehicle hazmat marker dataset called VisInt-VHM and introduces a compact hazmat marker detection network called HMD-Net. Experimental results demonstrate that HMD-Net achieves a better trade-off between detection accuracy and inference speed.
Article
Computer Science, Artificial Intelligence
Jing Zhang, Qingyong Li, YangLi-ao Geng, Wen Wang, Wenju Sun, Chuan Shi, Zhengming Ding
Summary: The paper introduces a Cluster-Prototype Matching (CPM) framework that utilizes the distribution information of samples to correct the biased relationships between seen and unseen classes, improving the performance and effectiveness of zero-shot learning.
PATTERN RECOGNITION
(2022)
Article
Computer Science, Artificial Intelligence
Siyu Cao, Wen Wang, Jing Zhang, Min Zheng, Qingyong Li
Summary: This paper proposes a novel few-shot fine-grained image classification framework, DLG, which enhances the discriminative ability of local structures. Experimental results show that DLG outperforms the state-of-the-art methods on fine-grained datasets.
INTERNATIONAL JOURNAL OF MACHINE LEARNING AND CYBERNETICS
(2022)
Article
Engineering, Civil
Xiaobao Li, Qingyong Li, Wen Wang, Lijun Guo
Summary: This paper proposes a unified coarse-to-fine unsupervised person re-identification framework named MNSR, which improves accuracy through Mutual Normalized Sparse Representation model and probability-guided label prediction method. In the metric model learning stage, reliable labels are selected for training to prevent noise samples from affecting the learning process, and experimental results demonstrate the superior performance of the MNSR method.
IEEE TRANSACTIONS ON INTELLIGENT TRANSPORTATION SYSTEMS
(2022)
Article
Meteorology & Atmospheric Sciences
Jiangtao Li, Xingqin An, Qingyong Li, Chao Wang, Haomin Yu, Xinyuan Zhou, Yangli-ao Geng
Summary: This study used the XGBoost algorithm to optimize the PM2.5 and O3 concentrations in Beijing, and found that the algorithm can better capture the spatial and temporal variation patterns of pollutant concentrations. The analysis of features and contribution values provided insights into the optimization principle of air pollution algorithm models. The optimized spatial distribution of pollutant concentrations was closer to the observed distribution compared to the WRF-Chem simulation.
ATMOSPHERIC RESEARCH
(2022)
Article
Computer Science, Artificial Intelligence
Jing Zhang, YangLi-ao Geng, Wen Wang, Wenju Sun, Zhirong Yang, Qingyong Li
Summary: This paper investigates how to perform zero-shot learning with fewer seen samples. A Distribution and Gradient constrained Embedding Model (DGEM) is proposed to predict the visual prototypes for the given semantic vectors of seen classes. Experimental results show that DGEM outperforms other established methods when each seen class has only 1/5 sample(s).
KNOWLEDGE-BASED SYSTEMS
(2022)
Article
Computer Science, Artificial Intelligence
Xiaobao Li, Qingyong Li, Fengjiao Liang, Wen Wang
Summary: Unsupervised person re-identification aims to train a model using fully unlabeled training images. Most successful approaches combine clustering-based pseudo-label prediction with ReID model learning in an alternating fashion. However, some person images are prone to be assigned wrong pseudo-labels. To address this, we propose a Multi-granularity Pseudo-label Collaboration (MPC) method.
COMPUTER VISION AND IMAGE UNDERSTANDING
(2023)
Review
Meteorology & Atmospheric Sciences
Weitao Lyu, Dong Zheng, Yang Zhang, Wen Yao, Rubin Jiang, Shanfeng Yuan, Dongxia Liu, Fanchao Lyu, Baoyou Zhu, Gaopeng Lu, Qilin Zhang, Yongbo Tan, Xuejuan Wang, Yakun Liu, Shaodong Chen, Lyuwen Chen, Qingyong Li, Yijun Zhang
Summary: This paper reviews the research progress on atmospheric electricity achieved in China during 2019-22, focusing on lightning detection and location techniques, thunderstorm electricity, lightning forecasting methods and techniques, physical processes of lightning discharge, high energy emissions and effects of thunderstorms on the upper atmosphere, and the effect of aerosol on lightning.
ADVANCES IN ATMOSPHERIC SCIENCES
(2023)
Article
Geochemistry & Geophysics
Shuyi He, Qingyong Li, Yang Liu, Wen Wang
Summary: This paper proposes a self-supervised semantic segmentation framework for remote sensing images with limited labeled data. It uses image inpainting as a pretext task and employs an adversarial training scheme to adaptively mask and restore local regions. Experimental results show that the method outperforms other approaches.
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS
(2022)
Proceedings Paper
Computer Science, Artificial Intelligence
Dong Li, Haomin Yu, Yangli-ao Geng, Xiaobao Li, Qingyong Li
Summary: This paper proposes a dual-stage dynamic spatio-temporal graph network (DDGNet) for PM2.5 prediction in different cities, which models dynamic correlations by dynamically constructing graphs, achieving state-of-the-art performance.
2021 IEEE INTERNATIONAL CONFERENCE ON BIG DATA (BIG DATA)
(2021)
Proceedings Paper
Computer Science, Artificial Intelligence
Siqi Cai, Wenyuan Xue, Qingyong Li, Peng Zhao
Summary: This study proposes a hybrid CTC-Attention decoder for Chinese text recognition based on the characteristic of Chinese word frequency distribution. Experimental results demonstrate the effectiveness of the proposed method, especially for long texts. The code will be publicly available on GitHub.
DOCUMENT ANALYSIS AND RECOGNITION, ICDAR 2021, PT III
(2021)
