检索结果-内蒙古大学图书馆

International Conference on Energy, Power, Environment, Control and Computing (ICEPECC 2025)

作者： Fatima N. Al-Aswadi Wafa’ Za’al Alma’aitah Huah Yong Chan Keng Hoon Gan Zahid Iqbal Ahlam Al-Dhamari Institute of Computer Science and Digital Innovation UCSI University Kuala Lumpur 56000 Malaysia Department of Intelligence Systems Faculty of Artificial Intelligence Al-Balqa Applied University Salt 19117 Jordan School of Computer Sciences Universiti Sains Malaysia 11800 Gelugor Pulau Pinang Malaysia Department of Computer Science Air University Kharian Pakistan Department of Electrical and Computer Engineering Curtin University Malaysia Miri 98009 Sarawak Malaysia

ISBN: (数字)9781837243150

Relation extraction and classification, collectively referred to as relation discovery, are fundamental tasks in the ontology construction process. In recent years, numerous studies have dedicated their efforts to devising hybrid approaches, supporting inclusive innovation in computational linguistics and ontology engineering. These approaches merge traditional methods with Machine Learning (ML) or Deep Learning (DL) techniques to tackle relation discovery tasks. This research examines the applicability and efficacy of four distinct models—Support Vector Machine (SVM), Naive Bayes (NB), k-Nearest Neighbors (KNN), and Convolutional Neural Network (CNN)—in relation extraction and classification. Two different datasets, ACL RD-TEC 2.0 and ACLRelAcS, are utilized, alongside two distinct text representation methodologies, BOW and TF-IDF-BOW. An in-depth analysis of the results, including the advantages and limitations of each model, is presented in the results and discussion sections.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hybrid Data-Free Knowledge Distillation 39

Hybrid Data-Free Knowledge Distillation

引用

39th Annual AAAI Conference on Artificial Intelligence, AAAI 2025

作者： Tang, Jialiang Chen, Shuo Gong, Chen School of Computer Science and Engineering Nanjing University of Science and Technology China Key Laboratory of Intelligent Perception and Systems for High-Dimensional Information of Ministry of Education China Jiangsu Key Laboratory of Image and Video Understanding for Social Security China Center for Advanced Intelligence Project RIKEN Japan Department of Automation Institute of Image Processing and Pattern Recognition Shanghai Jiao Tong University China

ISBN: (纸本)157735897X

Data-free knowledge distillation aims to learn a compact student network from a pre-trained large teacher network without using the original training data of the teacher network. Existing collection-based and generation-based methods train student networks by collecting massive real examples and generating synthetic examples, respectively. However, they inevitably become weak in practical scenarios due to the difficulties in gathering or emulating sufficient real-world data. To solve this problem, we propose a novel method called Hybrid Data-Free Distillation (HiDFD), which leverages only a small amount of collected data as well as generates sufficient examples for training student networks. Our HiDFD comprises two primary modules, i.e., the teacher-guided generation and student distillation. The teacher-guided generation module guides a Generative Adversarial Network (GAN) by the teacher network to produce high-quality synthetic examples from very few real-world collected examples. Specifically, we design a feature integration mechanism to prevent the GAN from overfitting and facilitate the reliable representation learning from the teacher network. Meanwhile, we drive a category frequency smoothing technique via the teacher network to balance the generative training of each category. In the student distillation module, we explore a data inflation strategy to properly utilize a blend of real and synthetic data to train the student network via a classifier-sharing-based feature alignment technique. Intensive experiments across multiple benchmarks demonstrate that our HiDFD can achieve state-of-the-art performance using 120 times less collected data than existing methods. Copyright © 2025, Association for the Advancement of Artificia Intelligence (***). All rights reserved.

关键词： Personnel training

来源：评论

学校读者我要写书评

暂无评论

Design of ANN Based Non-Linear Network Using Interconnection of Parallel Processor

引用

computer systems science & engineering 2023年第9期46卷 3491-3508页

作者： Anjani Kumar Singha Swaleha Zubair Areej Malibari Nitish Pathak Shabana Urooj Neelam Sharma Department of Computer Science Aligarh Muslim UniversityAligarh202002India Department of Industrial and Systems Engineering College of EngineeringPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia Department of Information Technology Bhagwan Parshuram Institute of Technology(BPIT)GGSIPUNew Delhi110078India Department of Electrical Engineering College of EngineeringPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia Department of Artificial Intelligence&Machine Learning Maharaja Agrasen Institute of Technology(MAIT)GGSIPUNew Delhi110086India

Suspicious mass traffic constantly evolves,making network behaviour tracing and structure more *** networks yield promising results by considering a sufficient number of processing elements with strong interconnections between *** offer efficient computational Hopfield neural networks models and optimization constraints used by undergoing a good amount of parallelism to yield optimal *** neural network(ANN)offers optimal solutions in classifying and clustering the various reels of data,and the results obtained purely depend on identifying a *** this research work,the design of optimized applications is presented in an organized *** addition,this research work examines theoretical approaches to achieving optimized results using *** mainly focuses on designing *** optimizing design approach of neural networks analyzes the internal process of the neural *** in developing the network are based on the interconnections among the hidden nodes and their learning *** methodology is proven best for nonlinear resource allocation problems with a suitable design and complex *** ANN proposed here considers more or less 46k nodes hidden inside 49 million connections employed on full-fledged parallel *** proposed ANN offered optimal results in real-world application problems,and the results were obtained using MATLAB.

关键词： Artificial neural network(ANN) multiprocessor hidden node nonlinear optimization parallel processing

来源：评论

学校读者我要写书评

暂无评论

Large Model for Small Data: Foundation Model for Cross-Modal RF Human Activity Recognition

arXiv

引用

arXiv 2024年

作者： Weng, Yuxuan Wu, Guoquan Zheng, Tianyue Yang, Yanbing Luo, Jun Department of Computer Science and Engineering Southern University of Science and Technology China Smart City Center Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology China College of Computer Science Sichuan University China College of Computing and Data Science Nanyang Technological University Singapore

Radio-Frequency (RF)-based Human Activity Recognition (HAR) rises as a promising solution for applications unamenable to techniques requiring computer visions. However, the scarcity of labeled RF data due to their non-interpretable nature poses a significant obstacle. Thanks to the recent breakthrough of foundation models (FMs), extracting deep semantic insights from unlabeled visual data become viable, yet these vision-based FMs fall short when applied to small RF datasets. To bridge this gap, we introduce FM-Fi, an innovative cross-modal framework engineered to translate the knowledge of vision-based FMs for enhancing RF-based HAR systems. FM-Fi involves a novel cross-modal contrastive knowledge distillation mechanism, enabling an RF encoder to inherit the interpretative power of FMs for achieving zero-shot learning. It also employs the intrinsic capabilities of FM and RF to remove extraneous features for better alignment between the two modalities. The framework is further refined through metric-based few-shot learning techniques, aiming to boost the performance for predefined HAR tasks. Comprehensive evaluations evidently indicate that FM-Fi rivals the effectiveness of vision-based methodologies, and the evaluation results provide empirical validation of FM-Fi’s generalizability across various environments. © 2024, CC BY.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

An Adaptive-Gain Recurrent Neural Network for Solving the Joint-Angle Drift Issues of Redundant Manipulators

An Adaptive-Gain Recurrent Neural Network for Solving the Jo...

引用

2023 International Annual Conference on Complex systems and Intelligent science, CSIS-IAC 2023

作者： Zheng, Boyu Li, Chunquan Zhang, Zhijun Yu, Junzhi Liu, P.X. School of Information Engineering Nanchang University Nanchang330031 China School of Automation Science and Engineering South China University of Technology Guangzhou510640 China College of Engineering Peking University Beijing100871 China Carleton University Department of Systems and Computer Engineering OttawaONK1S 5B6 Canada

ISBN: (纸本)9798350309003

This paper proposes a new adaptive-gain recurrent neural network (AG-RNN) to effectively cope with the joint-angle drift issues in redundant manipulators. Specifically, a joint-angle drift-free with the feedback control of the velocity layer motion equation (JADF-FC) is proposed via an optimization criterion for synchronously optimizing linear terms and quadratic. Then, the JADF-FC is reasonably formulated into a standard quadratic programming (QP) issue. Different from the previous recurrent neural networks (RNNs), the AG-RNN proposed in this paper constructs an error-based differential equation with a new adaptive-gain. It should be noted that the proposed adaptive-gain does not gradually approach infinity as time increases, which is more in line with actual hardware implementation requirements than the existing time-variant-gain. The adaptive-gain can reduce the joint-angle drift errors of the redundant manipulator. Therefore, the proposed AG-RNN can solve the QP problem of the manipulator more effectively and quickly. To validate the performance of the proposed AG-RNN, it is compared with representative RNNs. The experimental results indicate that smaller joint-angle drift errors can be get by the proposed AG-RNN solving JADF-FC scheme than the other solutions when solving the joint-angle drift issues. © 2023 IEEE.

关键词： Quadratic programming

来源：评论

学校读者我要写书评

暂无评论

Evaluating Lightning-Caused Wildfire Risk to Alaska’s Power Grid Infrastructure

Evaluating Lightning-Caused Wildfire Risk to Alaska’s Power...

引用

Annual Meeting Industrial and Commercial Power systems Technical Conference

作者： Kathleen DeMichele Masoumeh Heidari Kapourchali Christine Waigl Claire A. Richards Micah Hahn Long Zhao Payman Dehghanian Mohammad Heidari Kapourchali Department of Computer Science and Engineering University of Alaska Anchorage Anchorage AK International Arctic Research Center University of Alaska Fairbanks Fairbanks AK Department of Nursing and Systems Science Washington State University Spokane WA Institute for Circumpolar Health Studies University of Alaska Anchorage Anchorage AK Department of Electrical Engineering and Computer Science South Dakota School of Mines and Technology Rapid City SD Department of Electrical and Computer Engineering George Washington University Washington DC Department of Electrical Engineering University of Alaska Anchorage Anchorage AK

ISBN: (数字)9798331508081

ISBN: (纸本)9798331508098

Accurate prediction of wildfire risk is crucial for power grid decision makers, allowing them to identify areas of high risk and optimize resource allocation through strategic planning. This is particularly critical in regions like Alaska, where the risk of wildfires poses a significant threat to remote and difficult-to-access power infrastructures. In this study, we aggregate multiple sources of regional climate, topographic, lightning, and fire data to analyze the risk to the electricity grid posed by lightning-caused wildfires in Alaska. A focus area is chosen based on notable changes in lightning activity and the occurrences of lightning-induced fires, along with its proximity to densely populated areas and electric utility infrastructure. To identify key factors, we perform a significance analysis on widely recognized contributors to lightning-ignited fires and subsequently eliminate redundant features. The risk of fire occurrence (target variable) is calculated as a weighted average of two factors: the frequency of fires in the region and the distance to the nearest fire. Machine learning models, including logistic regression and multilayer perceptron, are applied. Ultimately, a decision tree regressor is used to enhance the interpretability of the results in predicting fire occurrence risk. Assigning risk values to the power grid in the designated focus area offers electric utilities in wildfire-prone regions valuable insights, enabling them to improve wildfire planning and mitigation strategies.

关键词： Wildfires Prevention and mitigation Merging Lightning Strategic planning Multilayer perceptrons Power grids Power industry Resource management Regression tree analysis

来源：评论

学校读者我要写书评

暂无评论

A Comparison of Multi Class Support Vector Machine vs Deep Convolutional Neural Network for Brain Tumor Classification

A Comparison of Multi Class Support Vector Machine vs Deep C...

引用

International Seminar on Application for Technology of Information and Communication (ISemantic)

作者： Nur Ryan Dwi Cahyo Christy Atika Sari Eko Hari Rachmawanto Cahaya Jatmoko Rabei Raad Ali Al-Jawry Mohamed Ayad Alkhafaji Department of Informatics Engineering University of Dian Nuswantoro Semarang Indonesia Study Program in Informatics Engineering University of Dian Nuswantoro Semarang Indonesia Department Computer Science Northern Technical University Semarang Indonesia Department of Computer Engineering Technology Northern Technical University Mosul Iraq Department of Computer Sciece National University of Science and Technology DhiQar Iraq

Multi SVM has long been one of the popular methods in classification, while DCNN has recently gained significant attention in image processing and pattern recognition. This research evaluates the effectiveness of Multi Class Support Vector Machine (M-SVM) and Deep Convolutional Neural Network (DCNN) techniques in classifying brain tumors. A dataset of 2660 3D medical images with dimensions 227 x 227 x 3; including Glioma, Meningioma, and Pituitary tumors, has been partitioned into distinct sets for both training and testing purposes. DCNN approach achieves excellent accuracy in identifying tumor names, with a training accuracy of 97.8% and 100% success rate in 9 experiments. The Multi SVM method demonstrates relatively good accuracy, with training accuracies ranging from 70% to 90% based on different kernel functions. These findings provide valuable insights for selecting appropriate methods in brain tumor classification and encourage further exploration of hybrid Multi SVM-DCNN approaches to enhance accuracy and reliability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Unsupervised Domain Adaptation for Action Recognition via Self-Ensembling and Conditional Embedding Alignment

Unsupervised Domain Adaptation for Action Recognition via Se...

引用

IEEE International Conference on Data Mining (ICDM)

作者： Indrajeet Ghosh Garvit Chugh Abu Zaher Md Faridee Nirmalya Roy Department of Information Systems University of Maryland Baltimore County United States Department of Computer Science and Engineering Indian Institute of Technology Jodhpur India Amazon United States

ISBN: (数字)9798331506681

ISBN: (纸本)9798331506698

Recent advancements in deep learning-based wearable human action recognition (wHAR) have improved the capture and classification of complex motions, but adoption remains limited due to the lack of expert annotations and domain discrepancies from user variations. Limited annotations hinder the model's ability to generalize to out-of-distribution samples. While data augmentation can improve generalizability, unsupervised augmentation techniques must be applied carefully to avoid introducing noise. Unsupervised domain adaptation (UDA) addresses domain discrepancies by aligning conditional distributions with labeled target samples, but vanilla pseudo-labeling can lead to error propagation. To address these challenges, we propose μDAR, a novel joint optimization architecture comprised of three functions: (i) consistency regularizer between augmented samples to improve model classification generalizability, (ii) temporal ensemble for robust pseudo-label generation and (iii) conditional distribution alignment to improve domain generalizability. The temporal ensemble works by aggregating predictions from past epochs to smooth out noisy pseudo-label predictions, which are then used in the conditional distribution alignment module to minimize kernel-based class-wise conditional maximum mean discrepancy (kCMMD) between the source and target feature space to learn a domain invariant embedding. The consistency-regularized augmentations ensure that multiple augmentations of the same sample share the same labels; this results in (a) strong generalization with limited source domain samples and (b) consistent pseudo-label generation in target samples. The novel integration of these three modules in μDAR results in a range of ~ 4-12% average macro-F1 score improvement over six state-of-the-art UDA methods in four benchmark wHAR datasets.

关键词： Adaptation models Annotations Noise Benchmark testing Data augmentation Noise measurement Human activity recognition Data mining Optimization

来源：评论

学校读者我要写书评

暂无评论

HYPERBOLIC DISTANCE-BASED SPEECH SEPARATION

arXiv

引用

arXiv 2024年

作者： Petermann, Darius Kim, Minje Indiana University Department of Intelligent Systems Engineering BloomingtonIN47408 United States University of Illinois at Urbana-Champaign Department of Computer Science IL61801 United States

In this work, we explore the task of hierarchical distance-based speech separation defined on a hyperbolic manifold. Based on the recent advent of audio-related tasks performed in non-Euclidean spaces, we propose to make use of the Poincaré ball to effectively unveil the inherent hierarchical structure found in complex speaker mixtures. We design two sets of experiments in which the distance-based parent sound classes, namely "near" and "far", can contain up to two or three speakers (i.e., children) each. We show that our hyperbolic approach is suitable for unveiling hierarchical structure from the problem definition, resulting in improved child-level separation. We further show that a clear correlation emerges between the notion of hyperbolic certainty (i.e., the distance to the ball's origin) and acoustic semantics such as speaker density, inter-source location, and microphone-to-speaker distance. © 2024, CC BY.

关键词： Source separation

来源：评论

学校读者我要写书评

暂无评论

Smart Home Gas Sensor Optimization Using Kalman Filter for Data Processing 9

Smart Home Gas Sensor Optimization Using Kalman Filter for D...

引用

9th International Conference on Informatics and Computing, ICIC 2024

作者： Somantri Thohir, Muhamad Ikhsan Sujjada, Alun Setiawan, Deni Komara, Ilham Efendi, Fikri Ardiansyah Muslih, Muhamad Abdullah, Zubaile Bin Saringat, Mohd Zainuri Department of Software Engineering University Tun Hussein Onn Malaysia Malaysia Nusa Putra University Informatics Engineering Study Program Sukabumi Indonesia Department of Information System Nusa Putra University Sukabumi Indonesia Universiti Tun Hussein Onn Malaysia Faculty of Computer Science and Information Technology Department of Software Engineering Johor Bahru Malaysia

ISBN: (纸本)9798331517601

The integration of Internet of Things (IoT) technologies into modern homes has enhanced safety and comfort, particularly in detecting gas leaks, which pose serious fire hazards. Gas leaks can often be detected by smell, but this method fails when no one is present. Previous research using microcontrollers and sensors to detect gas leaks faced challenges with accuracy due to noise interference. This paper proposes the use of a Kalman filter, developed by Rudolf Emil Kalman in 1960, to improve gas leak detection accuracy by filtering out noise. The system comprises an Arduino Nano, ESP8266 WiFi module, MQ-2 gas sensor, buzzer, and ThingSpeak cloud platform. By applying the Kalman filter, noise and data oscillations are reduced, enhancing detection accuracy. Experimental results show the system effectively detects gas leaks, provides real-time data, and triggers alarms. Future improvements could include additional sensors and features to further increase the system's reliability and functionality. © 2024 IEEE.

关键词： gas leak detection Internet of Things (IoT) Kalman filter MQ-2 sensor noise reduction

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：