检索结果-内蒙古大学图书馆

Label Distribution Learning with Label Correlations on Local Samples

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2021年第4期33卷 1619-1631页

作者： Jia, Xiuyi Li, Zechao Zheng, Xiang Li, Weiwei Huang, Sheng-Jun Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing 210014 Peoples R China Nanjing Univ Aeronaut & Astronaut Coll Astronaut Nanjing 210016 Peoples R China Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing 210016 Peoples R China

Label distribution learning (LDL) is proposed for solving the label ambiguity problem in recent years, which can be seen as an extension of multi-label learning. To improve the performance of label distribution learning, some existing algorithms exploit label correlations in a global manner that assumes the label correlations are shared by all instances. However, the instances in different groups may share different label correlations, and few label correlations are globally applicable in real-world tasks. In this paper, two novel label distribution learning algorithms are proposed by exploiting label correlations on local samples, which are called GD-LDL-SCL and Adam-LDL-SCL, respectively. To utilize the label correlations on local samples, the influence of local samples is encoded, and a local correlation vector is designed as the additional features for each instance, which is based on the different clustered local samples. Then, the label distribution for an unseen instance can be predicted by exploiting the original features and the additional features simultaneously. Extensive experiments on some real-world data sets validate that our proposed methods can address the label distribution problems effectively and outperform state-of-the-art methods.

关键词： Correlation clustering algorithms Silicon compounds Optimization Computer science Visualization Trees insulation Multi-label learning label distribution learning label correlations

来源：评论

学校读者我要写书评

暂无评论

Optimized Energy - Efficient Path Planning Strategy in WSN With Multiple Mobile Sinks

引用

IEEE ACCESS 2021年 9卷 82833-82847页

作者： Al-Kaseem, Bilal R. Taha, Zahraa K. Abdulmajeed, Sarah W. Al-Raweshidy, Hamed S. Al Iraqia Univ Coll Engn Dept Elect Engn Baghdad 7366 Iraq Al Iraqia Univ Coll Engn Dept Network Engn Baghdad 7366 Iraq Al Iraqia Univ Coll Engn Dept Comp Engn Baghdad 7366 Iraq Brunel Univ London Coll Engn Design & Phys Sci Wireless Networks & Commun Ctr WNCC Dept Elect & Comp Engn London UB8 3PH England

Recently, wireless sensor networks (WSNs) were perceived as the foundation infrastructure that paved the way to the emergence of the Internet of Things (IoT). However, a challenging issue exists when WSNs are integrated into the IoT because of high energy consumption in their nodes and poor network lifespan. Therefore, the elementary discussions in WSN are energy scarcity in sensor nodes, sensors' data exchange, and routing protocols. To address the aforesaid shortcomings, this paper develops an optimized energy-efficient path planning strategy that prolongs the network lifetime and enhances its connectivity. The proposed approach has four successive procedures: initially, the sensing field is partitioned into equal regions depending on the number of deployed mobile sinks that eliminate the energy-hole problem. A new heuristic clustering approach called stable election algorithm (SEA) is introduced to minimize the message exchange between sensor nodes and prevent frequent cluster heads rotation. A sojourn location determination algorithm is proposed based on the minimum weighted vertex cover problem (MWVCP) to find the best position for the sinks to stop and collect the data from cluster heads. Finally, three optimization techniques are utilized to evaluate the optimized mobile sinks' trajectories using multi-objective evolutionary algorithms (MOEAs). Whilst the performance of the developed work was evaluated in terms of cluster heads number, network lifetime, the execution time of the sinks' sojourn locations determination algorithm, the convergence rate of optimization techniques, and data delivery. The simulation scenarios conducted in MATLAB and the obtained results showed that the introduced approach outperformed comparable existing schemes. It succeeded in prolonging the network lifetime up to 66% compared to existing routing protocols.

关键词： Wireless sensor networks Internet of Things Wireless communication Sensors clustering algorithms Batteries Routing protocols Ant colony optimization clustering protocol genetic algorithm IoT M2M multiple mobile sinks optimized path routing protocol simulated annealing stable election algorithm WSN

来源：评论

学校读者我要写书评

暂无评论

Random Forest Model for Trip End Identification Using Cellular Phone and Points of Interest Data

引用

TRANSPORTATION RESEARCH RECORD 2021年第7期2675卷 454-466页

作者： Yang, Fei Wang, Yanchen Jin, Peter J. Li, Dingbang Yao, Zhenxing Southwest Jiaotong Univ Sch Transportat & Logist Chengdu Sichuan Peoples R China Rutgers State Univ Dept Civil & Environm Engn Piscataway NJ USA Changan Univ Coll Transportat Engn Xian Peoples R China

Cellular phone data has been proven to be valuable in the analysis of residents' travel patterns. Existing studies mostly identify the trip ends through rule-based or clustering algorithms. These methods largely depend on subjective experience and users' communication behaviors. Moreover, limited by privacy policy, the accuracy of these methods is difficult to assess. In this paper, points of interest data is applied to supplement cellular phone data's missing information generated by users' behaviors. Specifically, a random forest model for trip end identification is proposed using multi-dimensional attributes. A field data acquisition test is designed and conducted with communication operators to implement synchronized cellular phone data and real trip information collection. The proposed identification approach is empirically evaluated with real trip information. Results show that the overall trip end detection precision and recall reach 95.2% and 88.7% with an average distance error of 269 m, and the time errors of the trip ends are less than 10 min. Compared with the rule-based approach, clustering algorithm, naive Bayes method, and support vector machine, the proposed method has better performance in accuracy and consistency.

关键词： clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

A Permutation-Based Model for Analysis of Resource Allocation Overheads in Vehicular Ad Hoc Networks

引用

IEEE ACCESS 2021年 9卷 12282-12290页

作者： Zhang, Xin Shen, Zhongyi Yang, Dacheng Beijing Univ Posts & Telecommun Sch Informat & Commun Engn Beijing 100876 Peoples R China

This paper considers the theoretical limits of resource allocation overheads in vehicular ad hoc networks (VANETs), where the resource allocation procedure may consume extensive system resources. A permutation-based model is proposed to analyze the overhead limits where the unordered resource allocation patterns are considered as a permutation process. The minimum bits to represent the resource allocation information can be derived based on the permutation-based model. The bipartite graph is applied to describe the resource reuse in the respective domains for practical scenarios, and the dynamicity is taken into consideration. The proposed permutation-based model is applied to analyze the resource allocation overheads of two typical resource allocation schemes in VANETs: time-division multiple access (TDMA) and cluster-based algorithms. The proposed methodology provides new insights in a tradeoff analysis between overheads and system efficiency when analyzing the capacity of the VANETs.

关键词： Resource management Vehicular ad hoc networks Analytical models clustering algorithms Protocols Topology Time division multiple access Overhead resource allocation vehicular ad hoc networks theoretical limit

来源：评论

学校读者我要写书评

暂无评论

Feature Selection with Data Field

引用

Chinese Journal of Electronics 2025年第4期23卷 661-665页

作者： Hanning Yuan Shuliang Wang Ying Li Jinghua Fan School of Software Beijing Institute of Technology Beijing China State Key Laboratory of Information Engineering in Surveying Mapping and Remote Sensing Wuhan China International School of Software Wuhan University Wuhan China

A new feature selection method is proposed for high-dimensional data clustering on the basis of data field. With the potential entropy to evaluate the importance of feature subsets, features are filtered by removing unimportant features or noises from the original datasets. Experiments show that the proposed method can sharply reduce the number of dimensions and effectively improve the clustering performance on WDBC dataset.

关键词： Electric potential High dimensional data Noise Education clustering algorithms Medical services Feature extraction Entropy Noise measurement

来源：评论

学校读者我要写书评

暂无评论

Machine-Learning-Aided Optical Fiber Communication System

引用

IEEE NETWORK 2021年第4期35卷 136-142页

作者： Pan, Xiaolong Wang, Xishuo Tian, Bo Wang, Chuxuan Zhang, Hongxin Guizani, Mohsen Beijing Univ Posts & Telecommun Beijing Peoples R China Qatar Univ Comp Sci & Engn Doha Qatar

The fiber optical network offers high speed, large bandwidth, and a high degree of reliability. However, the development of optical communication technology has hit a bottleneck due to several challenges such as energy loss, cost, and system capacity approaching the Shannon limit. As a powerful tool, machine learning technology provides a strong driving force for the development of various industries and greatly promotes the development of society. Machine learning also provides a new possible solution to achieve greater transmission capacities and longer transmission distances in optical communications. In this article, we introduce the application of machine learning in optical communication network systems. Three use cases are presented to evaluate the feasibility of our proposed architecture. In the transmission layer, the principal-component-based phase estimation algorithm is used for phase noise recovery in coherent optical systems, and the K-means algorithm is adopted to reduce the influence of nonlinear noise in probabilistic shaping systems. As for the network layer, the long short-term memory algorithm and the genetic algorithm are suitable for making traffic predictions and determining reasonable placement locations of remote radio heads in centralized radio access networks. Extensive simulations and experiments are conducted to evaluate the proposed algorithm in comparison to the state-of-the-art schemes. The results show the performance of three use cases. Machine learning algorithms applied to the transmission layer can greatly promote the performance of digital signal processing without increasing the complexity. Machine learning algorithms applied to the network layer can provide a more appropriate channel allocation plan in the era of high-speed communication. Ultimately, the intent of this article is to serve as a basis for stimulating more research in machine learning in optical communications.

关键词： Machine learning algorithms Signal processing algorithms clustering algorithms Machine learning Optical fiber networks Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

AirNN: A Featherweight Framework for Dynamic Input-Dependent Approximation of CNNs

引用

IEEE TRANSACTIONS ON COMPUTER-AIDED DESIGN OF INTEGRATED CIRCUITS AND SYSTEMS 2021年第10期40卷 2090-2103页

作者： Hemmat, Maedeh San Miguel, Joshua Davoodi, Azadeh Univ Wisconsin Dept Elect & Comp Engn 1415 Johnson Dr Madison WI 53706 USA

In this work, we propose AirNN, a novel framework which enables dynamic approximation of an already-trained convolutional neural network (CNN) in hardware during inference. AirNN enables input-dependent approximation of the CNN to achieve energy saving without much degradation in its classification accuracy at runtime. For each input, AirNN uses only a fraction of the CNN's weights based on that input (with the rest remaining 0) to conduct the inference. Consequently, energy saving is possible due to fewer number of fetches from off-chip memory as well as fewer multiplications for majority of the inputs. To achieve per-input approximation, we propose a clustering algorithm that groups similar weights in the CNN based on their importance, and design an iterative framework that decides dynamically how many clusters of weights should be fetched from off-chip memory for each individual input. We also propose new hardware structures to implement our framework on top of a recently proposed FPGA-based CNN accelerator. In our experiments with popular CNNs, we, on average, show 49% energy saving with less than 3% degradation in classification accuracy due to doing inference with only a fraction of the weights for the majority of the inputs. We also propose a greedy interleaving scheme, implemented in hardware, in order to improve the performance of the iterative procedure and compensate for its latency overhead.

关键词： Hardware clustering algorithms Approximation algorithms Neurons Degradation Neural networks Inference algorithms Approximate computation early termination input-dependent approximation low-energy design neural network accelerators

来源：评论

学校读者我要写书评

暂无评论

Efficient and Accurate Selection of Optimal Collective Communication algorithms Using Analytical Performance Modeling

引用

IEEE ACCESS 2021年 9卷 109355-109373页

作者： Nuriyev, Emin Lastovetsky, Alexey Univ Coll Dublin Sch Comp Sci Dublin D04 V1W8 4 Ireland

The performance of collective operations has been a critical issue since the advent of Message Passing Interface (MPI). Many algorithms have been proposed for each MPI collective operation but none of them proved optimal in all situations. Different algorithms demonstrate superior performance depending on the platform, the message size, the number of processes, etc. MPI implementations perform the selection of the collective algorithm empirically, executing a simple runtime decision function. While efficient, this approach does not guarantee the optimal selection. As a more accurate but equally efficient alternative, the use of analytical performance models of collective algorithms for the selection process was proposed and studied. Unfortunately, the previous attempts in this direction have not been successful. We revisit the analytical model-based approach and propose two innovations that significantly improve the selective accuracy of analytical models: (1) We derive analytical models from the code implementing the algorithms rather than from their high-level mathematical definitions. This results in more detailed and relevant models. (2) We estimate model parameters separately for each collective algorithm and include the execution of this algorithm in the corresponding communication experiment. We experimentally demonstrate the accuracy and efficiency of our approach using Open MPI broadcast and gather algorithms and two different Grid'5000 clusters and one supercomputer.

关键词： Analytical models Mathematical model clustering algorithms Technological innovation Message passing Standards Solid modeling Message passing collective communication algorithms communication performance modeling MPI

来源：评论

学校读者我要写书评

暂无评论

Machine Learning at the Edge: A Data-Driven Architecture With Applications to 5G Cellular Networks

引用

IEEE TRANSACTIONS ON MOBILE COMPUTING 2021年第12期20卷 3367-3382页

作者： Polese, Michele Jana, Rittwik Kounev, Velin Zhang, Ke Deb, Supratim Zorzi, Michele Univ Padua Dept Informat Engn DEI I-35131 Padua Italy Northeastern Univ Inst Wireless Internet Things Boston MA 02120 USA AT&T Labs Bedminster NJ 07921 USA Dataminr New York NY USA Facebook New York NY USA

The fifth generation of cellular networks (5G) will rely on edge cloud deployments to satisfy the ultra-low latency demand of future applications. In this paper, we argue that such deployments can also be used to enable advanced data-driven and Machine Learning (ML) applications in mobile networks. We propose an edge-controller-based architecture for cellular networks and evaluate its performance with real data from hundreds of base stations of a major U.S. operator. In this regard, we will provide insights on how to dynamically cluster and associate base stations and controllers, according to the global mobility patterns of the users. Then, we will describe how the controllers can be used to run ML algorithms to predict the number of users in each base station, and a use case in which these predictions are exploited by a higher-layer application to route vehicular traffic according to network Key Performance Indicators (KPIs). We show that the prediction accuracy improves when based on machine learning algorithms that rely on the controllers' view and, consequently, on the spatial correlation introduced by the user mobility, with respect to when the prediction is based only on the local data of each single base station.

关键词： 5G mobile communication Cellular networks Base stations Machine learning algorithms Machine learning Computer architecture clustering algorithms 5G machine learning edge controller prediction mobility big data

来源：评论

学校读者我要写书评

暂无评论

A clustering and weighted euclidean norm-based short-term load forecasting of normal and anomalous days

A clustering and weighted euclidean norm-based short-term lo...

引用

International Conference on Electrical and Electronics Engineering, ICEEE 2020

作者： Babita Jain, M. Rao, Venu Gopala Poornima College of Engineering JaipurRajasthan India Mody University LaxmangarhRajasthan India

ISBN: (纸本)9789811546914

The paper presents a clustering and the weighted Euclidean distance measure-based combined approach for short-term load forecasting (STLF) of a day ahead hourly load of normal and anomalous days. Hourly load, temperature, humidity and day type are selected as the key variables of the data-set used. The input data-set for the clustering algorithm follows distinct day-type similarity criterion for normal and anomalous days. The model uses a two-tier architecture. In tier one, the input data-set is grouped into a pre-defined number of cluster subsets using weather parameter similarity. In tier two, the forecast day cluster subset is narrowed down to obtain five similar days of the forecast day using weighted Euclidean distance norm. The hourly loads of the similar days are then averaged to obtain the actual hourly load forecasts. The technique is tested for the eleven electrical load zones of the New York Independent System Operator (NYISO) over a period of three years. The performance evaluation of the proposed approach is done in comparison with the conventional neural network-based model implemented by the NYISO. Mean absolute percentage error (MAPE) is taken as the performance measure. The proposed technique gives lower MAPE values for all day types, seasons and terrains. © The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd 2021.

关键词： clustering algorithms

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：