检索结果-内蒙古大学图书馆

International Conference on Computational Intelligence and Applications (ICCIA)

作者： Gaoxiang Cong Jianxiong Wan Tianyang Hua Jie Zhou Hongxun Niu School of Information Engineering Inner Mongolia University of Technology Hohhot China Inner Mongolia Autonomous Region Engineering & Technology Research Center of Big Data Based Software Service School of Data Science and Application Inner Mongolia University of Technology Hohhot China School of Data Science and Application Inner Mongolia University of Technology Hohhot China

ISBN: (数字)9781728160429

ISBN: (纸本)9781728160436

data Centers (DC) requires massive monitoring for thermal and energy efficiency. Currently, popular wireless DC monitoring solutions include Zigbee and Bluetooth, etc. However, these solutions are typically short-range wireless communication technologies, leading to serious scalability issues. In this paper, we design and implement a wireless DC thermal monitoring system based on LoRa (Long Range). The system consists of data Acquisition Subsystem (DAS), data Transmission Subsystem (DTS), and Backend Monitoring Subsystem (BMS), where the thermal data are collected via LoRa network with star topology and routed to the BMS for thermal monitoring and fault diagnosis. An advantage of our solution is that the number of nodes that is necessary to cover the data center is significantly reduced due to the long-range communication of LoRa technology. In addition, we further cut the energy consumption of the system by a customized design of the end device such that all irrelevant peripheral components are removed. Finally, we show how dependable and real-time DC thermal monitoring can be achieved by using our solution in a field deployment.

关键词： Logic gates Monitoring Humidity Temperature distribution Wireless communication Servers Wireless sensor networks

来源：评论

学校读者我要写书评

暂无评论

A codebook-driven approach for low-light image enhancement

引用

engineering Applications of Artificial Intelligence 2025年 156卷

作者： Xu Wu Xianxu Hou Zhihui Lai Jie Zhou Ya-nan Zhang Witold Pedrycz Linlin Shen College of Computer Science and Software Engineering Shenzhen University Shenzhen 518060 China School of AI and Advanced Computing Xi’an Jiaotong-Liverpool University Suzhou 215400 China National Engineering Laboratory for Big Data System Computing Technology Shenzhen University Shenzhen 518060 China School of Computer Science Sichuan Normal University Chengdu 610065 China The Department of Electrical & Computer Engineering University of Alberta Alberta T6G 1H9 Canada

Low-light image enhancement (LLIE) aims to improve low-illumination images. However, existing methods face two challenges: (1) uncertainty in restoration from diverse brightness degradations; (2) loss of texture and color information caused by noise suppression and light enhancement. In this paper, we propose a novel enhancement approach, CodeEnhance, by leveraging discrete codebook priors and image refinement to address these challenges. In particular, we reframe LLIE as learning an image-to-code mapping from low-light images to discrete codebook, which has been learned from high-quality images. To enhance this process, a Semantic Embedding Module (SEM) is introduced to integrate semantic information with low-level features, and a Codebook Shift (CS) mechanism, designed to adapt the pre-learned codebook to better suit the distinct characteristics of our low-light dataset. Additionally, we present an Interactive Feature Transformation (IFT) module to refine texture and color information during image reconstruction, allowing for interactive enhancement based on user preferences. Extensive experiments on both real-world and synthetic benchmarks demonstrate that the incorporation of prior knowledge and controllable information transfer significantly enhances LLIE performance in terms of quality and fidelity. The proposed CodeEnhance exhibits superior robustness to various degradations, including uneven illumination, noise, and color distortion. The code can be obtained from https://***/csxuwu/CodeEnhance or https://***/*** .

关键词： Image coding

来源：评论

学校读者我要写书评

暂无评论

Cross-domain Object Detection through Coarse-to-Fine Feature Adaptation

Cross-domain Object Detection through Coarse-to-Fine Feature...

引用

Conference on Computer Vision and Pattern Recognition (CVPR)

作者： Yangtao Zheng Di Huang Songtao Liu Yunhong Wang Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University State Key Laboratory of Software Development Environment Beihang University School of Computer Science and Engineering Beihang University Beijing China

ISBN: (数字)9781728171685

ISBN: (纸本)9781728171692

Recent years have witnessed great progress in deep learning based object detection. However, due to the domain shift problem, applying off-the-shelf detectors to an unseen domain leads to significant performance drop. To address such an issue, this paper proposes a novel coarse-to-fine feature adaptation approach to cross-domain object detection. At the coarse-grained stage, different from the rough image-level or instance-level feature alignment used in the literature, foreground regions are extracted by adopting the attention mechanism, and aligned according to their marginal distributions via multi-layer adversarial learning in the common feature space. At the fine-grained stage, we conduct conditional distribution alignment of foregrounds by minimizing the distance of global prototypes with the same category but from different domains. Thanks to this coarse-to-fine feature adaptation, domain knowledge in foreground regions can be effectively transferred. Extensive experiments are carried out in various cross-domain detection scenarios. The results are state-of-the-art, which demonstrate the broad applicability and effectiveness of the proposed approach.

关键词： Feature extraction Subspace constraints Object detection Detectors Task analysis Semantics Prototypes

来源：评论

学校读者我要写书评

暂无评论

DFPL: Decentralized Federated Prototype Learning Across Heterogeneous data Distributions

arXiv

引用

arXiv 2025年

作者： Zhang, Hongliang Xu, Fenghua Yu, Zhongyuan Hu, Chunqiang Pang, Shanchen Wang, Xiaofen Yu, Jiguo School of Computer Science and Technology Qilu University of Technology Shandong Jinan250353 China Cyber Security Institute University of Science and Technology of China Anhui Hefei230026 China School of Information Science and Engineering Lanzhou University Gansu Lanzhou730000 China School of Big Data and Software Engineering Chongqing University Chongqing400044 China Shandong Qingdao266580 China School of Computer Science and Engineering University of Electronic Science and Technology of China Sichuan Chengdu611731 China

Federated learning is a distributed machine learning paradigm that enables the collaborative training of multiple clients through centralized model aggregation. However, standard federated learning relies on a centralized server, making it vulnerable to server failures. While existing solutions utilize blockchain technology to implement Decentralized Federated Learning (DFL), the statistical heterogeneity of data distributions among clients severely degrades the DFL’s performance. Driven by this issue, this paper proposes a decentralized federated prototype learning framework, named DFPL, which significantly improves the performance of distributed machine learning across heterogeneous data distributions. Specifically, our framework introduces prototype learning into DFL to address statistical heterogeneity, which greatly reduces the number of parameters exchanged between clients. Additionally, blockchain is embedded into our framework, enabling the training and mining processes to be implemented at each client. From a theoretical perspective, we provide convergence guarantee of DFPL by combining resource allocation for training and mining. The experiments highlight the superiority of our DFPL framework in communication efficiency and test performance across three benchmark datasets with heterogeneous data distributions. © 2025, CC BY.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Single-path Bit Sharing for Automatic Loss-aware Model Compression

arXiv

引用

arXiv 2021年

作者： Liu, Jing Zhuang, Bohan Chen, Peng Shen, Chunhua Cai, Jianfei Tan, Mingkui The School of Software Engineering South China University of Technology China The Key Laboratory of Big Data and Intelligent Robot South China University of Technology Ministry of Education China The Pazhou Laboratory Guangzhou China The School of Computer Science the University of Adelaide Australia The Faculty of Information Technology Monash University Australia The College of Computer Science and Technology Zhejiang University Hangzhou China

Network pruning and quantization are proven to be effective ways for deep model compression. To obtain a highly compact model, most methods first perform network pruning and then conduct quantization based on the pruned model. However, this strategy may ignore that the pruning and quantization would affect each other and thus performing them separately may lead to sub-optimal performance. To address this, performing pruning and quantization jointly is essential. Nevertheless, how to make a trade-off between pruning and quantization is non-trivial. Moreover, existing compression methods often rely on some pre-defined compression configurations (i.e., pruning rates or bitwidths). Some attempts have been made to search for optimal configurations, which however may take unbearable optimization cost. To address these issues, we devise a simple yet effective method named Single-path Bit Sharing (SBS) for automatic loss-aware model compression. To this end, we consider the network pruning as a special case of quantization and provide a unified view for model pruning and quantization. We then introduce a single-path model to encode all candidate compression configurations, where a high bitwidth value will be decomposed into the sum of a lowest bitwidth value and a series of re-assignment offsets. Relying on the single-path model, we introduce learnable binary gates to encode the choice of configurations and learn the binary gates and model parameters jointly. More importantly, the configuration search problem can be transformed into a subset selection problem, which helps to significantly reduce the optimization difficulty and computation cost. In this way, the compression configurations of each layer and the trade-off between pruning and quantization can be automatically determined. Extensive experiments on CIFAR-100 and ImageNet show that SBS significantly reduces computation cost while achieving promising performance. For example, our SBS compressed MobileNetV2 achieves

关键词： Encoding (symbols)

来源：评论

学校读者我要写书评

暂无评论

A Multi-Objective Natural Aggregation Approach Based Residential Energy Management with Photovoltaic Power Penetration

A Multi-Objective Natural Aggregation Approach Based Residen...

引用

Power and Electrical engineering (ACPEE), 5th Asia Conference on

作者： Zehua Zhao Fengji Luo Yanli Liu Ali Dorri Min Gao School of Big Data & Software Engineering Chongqing University Chongqing China School of Civil Engineering The University of Sydney Sydney Australia School of Electrical & Information Engineering Tianjin University Tianjin China School of Computer Science & Engineering The University of New South Wales Sydney Australia

ISBN: (数字)9781728152813

ISBN: (纸本)9781728152820

Residential energy management (REM) has drawn wide attention in demand side management research recent years. This paper proposed a multi-objective optimization approach for REM in a high photovoltaic solar power penetrated environment. The proposed model manages three kinds of residential energy resources (RERs): a rooftop photovoltaic source, a heating, ventilation and air conditioning (HVAC) system, and multiple controllable household appliances. The proposed model aims at achieving following three objectives: (1) to minimize the home electricity cost; (2) to maximize the indoor thermal comfort for the user; and (3) to maximize the usage convenience of non-thermal appliances. An adaptive thermal comfort model is employed to model the indoor thermal comfort degree of the user, and a user disturbance metric is proposed to measure the psychological disturbance of the household appliance operation on the user's preference. A new multi-objective optimization solver, i.e., MultiObjective Natural Aggregation Algorithm (MONAA), is proposed to find the approximated Pareto frontier of the model. Simulations are conducted to validate the proposed method.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Detecting Potential Serious Adverse Drug Reactions Using Sequential Pattern Mining Method 9

Detecting Potential Serious Adverse Drug Reactions Using Seq...

引用

9th IEEE International Conference on software engineering and Service Science, ICSESS 2018

作者： Yuwen, Lu Chen, Shuyu Zhang, Hancui School of Big Data and Software Engineering Chongqing University Chongqing China

ISBN: (纸本)9781538665640

Adverse drug reactions (ADRs), defined as noxious and unintended responses to drugs at normal doses, cause hospitalization and death all around the world, and cost billions of dollars every year. Different techniques have been explored to discover potential ADRs using materials from various sources, such as spontaneous reporting systems, electronic medical records and clinical related social media data. In this paper, we proposed a framework that combines sequential data mining method Prefix-Span and classic disproportionality based method Proportional Report Ratio (PRR) to detect potential serious ADRs, taking temporal information among clinical events into consideration. And Naranjo algorithm, a questionnaire designed to determine the likelihood of causal relationships between drugs and reactions, is introduced to verify the validity of the results we obtain. In the end, we got 127 pair-wise frequent patterns of drugs and reactions, and 52% of the drugs overlap with that of the Naranjo algorithm. And the comparison of those two approaches shows that detecting potential ADRs based on disease category may increase the accuracy and creditability of the results. © 2018 IEEE.

关键词： data mining

来源：评论

学校读者我要写书评

暂无评论

An HVSM-based GRU approach to predict cross-version software defects

引用

International Journal of Performability engineering 2020年第6期16卷 979-990页

作者： Bai, Xue Zhou, Hua Yang, Hongji School of Software Yunnan University Kunming650000 China College of Big Data and Intelligent Engineering Southwest Forestry University Kunming650000 China School of Informatics Leicester University LeicesterLE1 7RH United Kingdom

Cross-version software Defect Prediction (CSDP) can be used to predict defect-prone modules or the number of defects in the latest version by using defect rules learned from historical versions. This technique can not only reduce the cost of iterative development of software and improve the reliability of new version, but also be used to understand the causes of defects and improve the software development process. Recently, much effort has been paid to build accurate cross-version defect prediction models, including quality defect predictors, and develop modeling techniques. Liu et al. proposed a new way to construct the predictor based on software code metrics and software process metrics. This new predictor is called Historical Version Sequence of Metrics (HVSM), which is processed with RNN. However, their model has some shortcomings, such as the missing of evolution information in HVSM, difficulty in training RNN, and lack of the ability to predict the number of defects, etc. To solve these problems, we add software evolution metrics to HVSM and bring in a new deep learning technique, Gate Recurrent Unit (GRU), to enhance the HVSM. Our HVSM-based GRU model can predict both defect-prone modules and the number of defects. The experimental results show that the HVSM with incremental evolution metrics provide better performance and in most cases the HVSM-based GRU model outperforms the commonly used baseline models in CSDP. © 2020 Totem Publisher, Inc. All rights reserved.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

FeatureBand: A feature selection method by combining early stopping and genetic local search 3rd

FeatureBand: A feature selection method by combining early s...

引用

3rd APWeb and WAIM Joint Conference on Web and big data, APWeb-WAIM 2019

作者： Xue, Huanran Jiang, Jiawei Shao, Yingxia Cui, Bin Center for Data Science National Engineering Laboratory for Big Data Analysis and Applications Peking University Beijing China Peking University Beijing China Tencent Inc. Shenzhen China Beijing Key Lab of Intelligent Telecommunications Software and Multimedia BUPT Beijing China

ISBN: (纸本)9783030260743

Feature selection is an important problem in machine learning and data mining. In reality, the wrapper methods are broadly used in feature selection. It treats feature selection as a search problem using a predictor as a black-box. However, most wrapper methods are time-consuming due to the large search space. In this paper, we propose a novel wrapper method, called FeatureBand, for feature selection. We use the early stopping strategy to terminate bad candidate feature subsets and avoid wasting of training time. Further, we use a genetic local search to generate new subsets based on previous ones. These two techniques are combined under an iterative framework in which we gradually allocate more resources for more promising candidate feature subsets. The experimental result shows that FeatureBand achieves a better trade-off between search time and search accuracy. It is 1.45 × to 17.6 × faster than the state-of-the-art wrapper-based methods without accuracy loss. © 2019, Springer Nature Switzerland AG.

关键词： Feature Selection

来源：评论

学校读者我要写书评

暂无评论

Signal approximation with Pascal’s triangle and sampling

Signal approximation with Pascal’s triangle and sampling

引用

Chinese Control and Decision Conference, CCDC

作者： Lei Chen Xinghuo Yu Jinhu Lü School of Automation Science and Electrical Engineering State Key Laboratory of Software Development Environment Beijing Advanced Innovation Center for Big Data and Brain Computing Beihang University Beijing China School of Engineering RMIT University Melbourne VIC Australia

ISBN: (数字)9781728158556

ISBN: (纸本)9781728158563

This brief explores the approximation properties of a unique basis expansion based on Pascal's triangle, which realizes a sampled-data driven approach between a continuous-time signal and its discrete-time representation. The roles of certain parameters, such as sampling time interval or model order, and signal characteristics, i.e., its curvature, on the approximation are investigated. Approximate errors in one and multiple-step predictions are analyzed. Furthermore, time-variant approximations under the thresholds of signal curvature are employed to narrow errors and provide flexibilities.

关键词： Shape Electronic mail Analytical models Mathematical analysis data models Upper bound

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：