检索结果-内蒙古大学图书馆

Short-Term Load Forecasting Based on Graph Convolution and Dendritic Deep Learning

IEEE Transactions on Network Science and Engineering 2025年

作者： Zhang, Chunyang Yu, Yang Zhang, Tengfei Song, keyu Wang, Yirui Gao, Shangce Nanjing University of Posts and Telecommunications College of Automation & College of Artificial Intelligence Nanjing210023 China Ningbo University Faculty of Electrical Engineering and Computer Science Zhejiang Key Laboratory of Mobile Network Application Technology Zhejiang315211 China University of Toyama Faculty of Engineering Toyama-shi930-8555 Japan

Short-term load forecasting (STLF) is a significant task to the planning, operation and control of future power systems. The increasing number of devices connected to the system has led to more complex characteristics and forms of load, which has brought considerable difficulties to the relevant methods in achieving higher load prediction accuracy and reliability. In this regard, this study proposes a deep learning model that combines graph convolutional network (GCN), gated recurrent unit (GRU), and dendritic neural model (DNM) to forecast electric load more accurately. Firstly, the sample load data is constructed into graph data with individual time steps as nodes. A GCN is used to extract the hidden features while allowing a full communication between the time steps feature data. A GRU is then used to capture the time-dependent relationship of the data. Finally, a dendritic layer instead of a fully connected layer is used as the output to integrate data features in depth. Experiments are conducted to verify the validity of the proposed model and compared it with several effective deep learning models, including CNN_LSTM, Transformer and Kolmogorov-Arnold Networks (KAN). The results show a significant improvement in prediction compared to the baseline models, with mean absolute percentage error(MAPE) of 1.62% and 3.98%, coefficient of determination(R2) of 0.983 and 0.928 respectively on two load datasets at different levels of aggregation, nationally and regionally. © 2013 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

t-PSI: Efficient Multi-party Private Set Intersection with Threshold 19

t-PSI: Efficient Multi-party Private Set Intersection with T...

引用

2022 IEEE SmartWorld, 19th IEEE International Conference on Ubiquitous Intelligence and Computing, 2022 IEEE International Conference on Autonomous and Trusted Vehicles Conference, 22nd IEEE International Conference on Scalable Computing and Communications, 2022 IEEE International Conference on Digital Twin, 8th IEEE International Conference on Privacy Computing and 2022 IEEE International Conference on Metaverse, SmartWorld/UIC/ATC/ScalCom/DigitalTwin/PriComp/Metaverse 2022

作者： Meng, Dan Fu, Zhihui Kong, Chao Qi, Yue Cao, Guitao Oppo Research Institute Shenzhen China Anhui Polytechnic University School of Computer and Information Wuhu China East China Normal University Shanghai Key Laboratory of Trustworthy Computing Shanghai China MoE Engineering Research Center of SH/HW Co-design Technology and Application Ecnu Shanghai China

ISBN: (纸本)9798350346558

Multi-party privacy set intersection (MPSI) enables multiple parties to compute the intersection of their datasets without leaking data privacy. Among existing MPSI protocols, e.g., KMPRT-based protocols, oblivious evaluation of programmable pseudo-random function (OPPRF) or oblivious pseudo-random function (OPRF) is typically used to generate pseudo random numbers, requiring frequent online interactions. To mitigate such communication overheads, we propose the threshold Privacy Set Intersection (t-PSI) protocol for multi-party, leveraging the Shamir Secret Sharing (SSS) protocol and the masking construction mechanism. We propose two types of masking construction mechanisms: Polynomial and Garbled Bloom-filter. To the best of our knowledge, this is the first Secret Sharing-based MPSI protocol. With the significant reduction of inter-party communication, the t-PSI protocol is highly scalable and can handle datasets ranging from 26 to 220. Further, with built-in fault tolerance, t-PSI works even if some parties went offline during protocol execution. As shown in the experiments, for five parties with datasets of 220 items each, the proposed t-PSI takes only 63 seconds, outperforming the state-of-the-art in [1]. © 2022 IEEE.

关键词： Fault tolerance

来源：评论

学校读者我要写书评

暂无评论

Full-atom protein pocket design via iterative refinement 23

Full-atom protein pocket design via iterative refinement

引用

Proceedings of the 37th International Conference on Neural Information Processing Systems

作者： Zaixi Zhang Zepu Lu Zhongkai Hao Marinka Zitnik Qi Liu Anhui Province Key Lab of Big Data Analysis and Application School of Computer Science and Technology University of Science and Technology of China and State Key Laboratory of Cognitive Intelligence Hefei Anhui China and Harvard University Anhui Province Key Lab of Big Data Analysis and Application School of Computer Science and Technology University of Science and Technology of China and State Key Laboratory of Cognitive Intelligence Hefei Anhui China Dept. of Comp. Sci. and Tech. Institute for AI THBI Lab BNRist Center Tsinghua-Bosch Joint ML Center Tsinghua Harvard University

The design and optimization of functional proteins that bind specific ligand molecules is paramount in therapeutics and bio-engineering. A critical yet formidable task in this endeavor is the design of the protein pocket, which is the cavity region of the protein where the ligand binds. Current methods are plagued by inefficient generation, inadequate context modeling of the ligand molecule, and the inability to generate side-chain atoms. Here, we present the Full-Atom Iterative Refinement (FAIR) method, designed to address these challenges by facilitating the co-design of protein pocket sequences, specifically residue types, and their corresponding 3D structures. FAIR operates in two steps, proceeding in a coarse-to-fine manner (transitioning from protein backbone to atoms, including side chains) for a full-atom generation. In each iteration, all residue types and structures are simultaneously updated, a process termed full-shot refinement. In the initial stage, the residue types and backbone coordinates are refined using a hierarchical context encoder, complemented by two structure refinement modules that capture both inter-residue and pocket-ligand interactions. The subsequent stage delves deeper, modeling the side-chain atoms of the pockets and updating residue types to ensure sequence-structure congruence. Concurrently, the structure of the binding ligand is refined across iterations to accommodate its inherent flexibility. Comprehensive experiments show that FAIR surpasses existing methods in designing superior pocket sequences and structures, producing average improvement exceeding 10% in AAR and RMSD metrics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Study on application of Artificial Intelligence in College English Teaching Management and Evaluation 3

A Study on Application of Artificial Intelligence in College...

引用

3rd International Conference on Electronic Informationtechnology and Smart Agriculture, ICEITSA 2023

作者： Feng, Tao Deng, Jianjie Liu, Yiwen Gao, Yanxia Peng, Ling Fu, Jinrong School of Foreign Languages Huaihua University Hunan Huaihua418000 China School of Computer and Artificial Intelligence Huaihua University Hunan Huaihua418000 China Key Laboratory of Wuling-Mountain Health Big Data Intelligent Processing and Application in Hunan Province Universities Hunan Huaihua418000 China Key Laboratory of Intelligent Control Technology for Wuling-Mountain Ecological Agriculture in Hunan Province Hunan Huaihua418000 China School of Computing Science University of Glasgow GlasgowG12 8QQ United Kingdom

ISBN: (纸本)9798400716775

With the rapid development of artificial intelligence, AI technology has been widely applied in various industries, including education. In English teaching, AI technology provides teachers with objective and instructive feedback, offers students a more personalized learning experience, helping them improve their English language skills, etc., and brings new development opportunities and challenges to English teaching as well. However, there are still some problems that need to be solved in college English teaching in our university. For this reason, based on the application of AI technology in college English teaching, this paper mainly explores the impact and application countermeasures of AI on college English teaching from the following four aspects: individuals with differences, teaching management, classroom teaching and teaching evaluation. In addition, the profound integration of college English education and AI will foster the advancement and transformation of college English instruction, ultimately serving as a catalyst for its reform. © 2023 ACM.

关键词： Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

A New Adaptive Image Sharpening Scheme

A New Adaptive Image Sharpening Scheme

引用

IEEE International Symposium on Information (IT) in Medicine and Education, ITME

作者： Tzong–Jer Chen Rong–Hui Lu School of Mathematics & Computer Science Wuyi University Wuyishan Fujian China Fujian Key Laboratory of Big Data Application and Intellectualization for Tea Industry Wuyi University Information Technology and Laboratory Management Center Wuyi University Wuyishan Fujian China

The sharpening is a scheme applied to highlight the intensity transitions in an image. This enhancement increases edge acutance and significantly improves the overall sharpness of processed images. Image sharpening is also of enormous importance in medicine for diagnosis. Noise overshoot and over-sharpening effects occur in traditional sharpening algorithms. The adaptive approach skips the noise producing a better sharpening result. Abrupt pixel value changes, if adequately emphasized, can be utilized effectively in the sharpening process. Since a residual image contains noise with edges, the bigger the filter size the greater the content of edges left in the residual image. Two residual images are produced first with different filter sizes and assumed a relationship between two residuals. A simple linear regression is measured using these residuals. This linear equation is an expectation and the deviation is the difference between two residuals. The window may contain edges if the residual value measured by larger size filter to some degree higher than that by smaller size filter. One standard deviation (±σ) is proposed as the threshold, in this work. A sensitive sharpening filter is adaptively then applied to those locations in image. It has been demonstrated that the proposed approach yields better results than the global sharpening filter on the measurements of Pratt's $P$ indices. A sharpening scheme without latent image noise amplification is useful for pattern recognition and machine-learning. In the future this scheme will be used to preprocess image segmentation and medical image.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3-D Near-Field Localization by Jointly Exploiting Spatial and Temporal Information Based on a Nonuniform Cross Array

3-D Near-Field Localization by Jointly Exploiting Spatial an...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zelong Yi Hua Chen Zhiwei Jiang Wei Liu Qing Wang Gang Wang Faculty of Electrical Engineering and Computer Science Ningbo University Ningbo China Zhejiang Key Laboratory of Mobile Network Application Technology Ningbo China School of Electronic Engineering and Computer Science Queen Mary University of London London UK School of Electronic Informatin Engineering Tianjin University Tianjin China

In this paper, an underdetermined three-dimensional (3-D) near-field source localization method is proposed, based on a two-dimensional (2-D) symmetric nonuniform cross array. Firstly, the fourth-order cumulant of the near-field observations with multiple delay lags is exploited to construct virtual far-field pseudo-observations, leading to increased degrees of freedom (DOF); then, 2-D angles of the nearfield sources are jointly estimated by employing the recently proposed sparse and parametric approach (SPA) and Vandermonde decomposition technique, eliminating the need for parameter discretization. To estimate the range term, the one-dimensional (1-D) MUSIC algorithm is applied by resorting to the conjugate symmetry property of the signal's autocorrelation function. Numerical results are provided to demonstrate the superiority of our method.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Residual and Attention Mechanism-Based Network for Aluminum-Silicon Alloy Microstructure Image Classification

Residual and Attention Mechanism-Based Network for Aluminum-...

引用

Information Systems and Computing technology (ISCTech), International Conference on

作者： Xiao Yang Tao Shen Jiancheng Yin Yingli Liu Faculty of Information Engineering and Automation Kunming University of Science and Technology Kunming China Yunnan Key Laboratory of Computer Technologies Application Kunming University of Science and Technology Kunming China Faculty of Materials Science and Engineering Kunming University of Science and Technology Kunming China

To enhance the accuracy of aluminum-silicon alloy microstructure image classification, we propose an aluminum-silicon alloy microstructure image classification model based on an improved residual network, incorporating attention mechanisms and multi-scale thinking. The model is built upon the ResNet18 architecture, utilizing enhanced basic residual blocks to select informative features. Additionally, a multi-scale attention module is incorporated into the network to fuse multi-scale features and further extract effective features. Experimental results on an aluminum-silicon alloy microstructure image classification dataset demonstrate that compared to the best-performing model currently available, the proposed model achieves an increase in classification accuracy, precision, recall, and F1 score by 3.17%, 0.74%, 5.53%, and 3.26%, respectively, reaching 94.97%, 96.39%, 93.97%, and 95.17%. These results indicate that the proposed model exhibits strong capabilities in classifying aluminum-silicon alloy microstructure images and can facilitate further research on the properties of aluminum-silicon alloys.

关键词：

来源：评论

学校读者我要写书评

暂无评论

When Adversarial Example Attacks Meet Vertical Federated Learning 19

When Adversarial Example Attacks Meet Vertical Federated Lea...

引用

ISBN: (纸本)9798350346558

Federated learning (FL) is proposed to enable efficient machine learning while protecting the privacy of user data. In non-federated scenarios, machine learning models are vulnerable to adversarial attacks. The attacker can construct elaborate adversarial examples, which can greatly reduce the performance of the victim model. It is worth exploring how FL models differ from traditional models in face of complex adversarial attacks. We are also curious about the robustness of different FL modeling methods against adversarial attacks, providing suggestions for selecting suitable FL models in practical applications. Consequently, we perform adversarial example attacks on vertical federated learning and design experiments to study how vertical federated models perform under adversarial example attacks. The simulation results show that federated models are more robust than non-federated models. Besides, in vertical federation learning, decision trees and XGBoost are much more robust than logistic regression and neural networks. © 2022 IEEE.

关键词： Decision trees

来源：评论

学校读者我要写书评

暂无评论

Expensive Multi-Objective Bayesian Optimization Based on Diffusion Models 39

Expensive Multi-Objective Bayesian Optimization Based on Dif...

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Li, Bingdong Di, Zixiang Lu, Yongfan Qian, Hong Wang, Feng Yang, Peng Tang, Ke Zhou, Aimin Shanghai Frontiers Science Center of Molecule Intelligent Syntheses Shanghai Institute of AI for Education School of Computer Science and Technology East China Normal University Shanghai 200062 China Key Laboratory of Advanced Theory and Application in Statistics and Data Science Ministry of Education China School of Computer Science Wuhan University Wuhan 430072 China Department of Statistics and Data Science Southern University of Science and Technology Shenzhen 518055 China Guangdong Provincial Key Laboratory of Brain-Inspired Intelligent Computation Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen 518055 China

ISBN: (纸本)157735897X

Multi-objective Bayesian optimization (MOBO) has shown promising performance on various expensive multi-objective optimization problems (EMOPs). However, effectively modeling complex distributions of the Pareto optimal solutions is difficult with limited function evaluations. Existing Pareto set learning algorithms may exhibit considerable instability in such expensive scenarios, leading to significant deviations between the obtained solution set and the Pareto set (PS). In this paper, we propose a novel Composite Diffusion Model based Pareto Set Learning algorithm (CDM-PSL) for expensive MOBO. CDM-PSL includes both unconditional and conditional diffusion model for generating high-quality samples efficiently. Besides, we introduce a weighting method based on information entropy to balance different objectives. This method is integrated with a guiding strategy to appropriately balancing different objectives during the optimization process. Experimental results on both synthetic and real-world problems demonstrates that CDM-PSL attains superior performance compared with state-of-the-art MOBO algorithms. Copyright © 2025, Association for the Advancement of Artificial Intelligence (***). All rights reserved.

关键词： Multiobjective optimization

来源：评论

学校读者我要写书评

暂无评论

A Novel Image Encryption Scheme Based on Poker Cross-Shuffling and Fractional Order Hyperchaotic System 10th

A Novel Image Encryption Scheme Based on Poker Cross-Shuffli...

引用

10th International Conference on computer Engineering and Networks, CENet 2020

作者： Chen, Zhong Zhao, Huihuang Chen, Junyao College of Computer Science and Technology Hengyang Normal University Hengyang421002 China Hunan Provincial Key Laboratory of Intelligent Information Processing and Application Hengyang421002 China Department of Computer Engineering AjouUniversity Suwon16499 Korea Republic of

ISBN: (纸本)9789811584619

A novel digital image encryption scheme based on Poker shuffling and fractional order hyperchaotic system is proposed. Poker intercrossing operation has nonlinearity and periodicity, and can form permutation group. Firstly, the plain image is transformed into block image, and each block is confused by Poker shuffling operation with key. Secondly, bit plane of every pixel in each block is confused. Lastly, the scrambled image is encrypted by fractional order hyperchaotic sequence. This encryption scheme provides a secure and efficient key stream, and guarantees a large key space. The experimental results and security analysis imply that the new encryption method has secure encryption effect and property of resisting common attacks. © 2021, The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Cryptography

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：