检索结果-内蒙古大学图书馆

SSRN 2023年

作者： He, Chunmei Zhou, Kang Tang, Jing Wu, Shengyu Ye, Zhengchun School of Computer Science School of Cyberspace Science Xiangtan University Hunan Xiangtan411105 China School of Mechanical Engineering Xiangtan University Hunan Xiangtan411105 China

The existing unsupervised domain adaption (UDA) methods achieve good domain adaptation results by minimizing domain differences based on measurement criteria or learning domain invariant features. However, these methods pay less attention to fine-grained aspects of processing, such as not focusing on hard samples or not fully considering the processing of hard samples in both source and target domains. To address these problems, unsupervised domain adaptation framework with Hard-samples Processing Strategy and Teacher-student Mechanism (HPSTM) is proposed in this paper. In the framework, we respectively define sample division criteria for both source and target domain and divide all samples into hard samples and easy samples. Then we further design systematic processing strategies of these samples. Specifically, we design a weighting strategy based on sample classification entropy for different samples in target domain, while we design the class prototypes to eliminate the hard samples for source domain. At last, we design the teacher-student guidance mechanism to explicitly learn discriminate features of hard samples and improve model`s generalization ability. The proposed framework can effectively promote the diversity learning of sample features and explicitly improve the discriminant performance of the model for hard samples. Extensive qualitative and quantitative experiments on five benchmark datasets verify that the proposed framework is effective and superior to most existing UDA methods. © 2023, The Authors. All rights reserved.

关键词： Image classification

来源：评论

学校读者我要写书评

暂无评论

High thermoelectric power factor in Ni-Fe alloy for active cooling applications

arXiv

引用

arXiv 2025年

作者： Li, Shuai Das, Sree Sourav Wang, Haobo Bati, Sujit Balachandran, Prasanna V. Shiomi, Junichiro Floro, Jerrold A. Zebarjadi, Mona Department of Materials Science and Engineering University of Virginia CharlottesvilleVA22904 United States Electrical and Computer Engineering Department University of Virginia CharlottesvilleVA22904 United States Department of Physics University of Virginia CharlottesvilleVA22904 United States Department of Mechanical Engineering University of Tokyo Tokyo Japan

Metallic thermoelectric materials are promising candidates for active cooling applications, where high thermal conductivity and a high thermoelectric power factor are essential to maximize effective thermal conductivity. While metals inherently possess high thermal and electrical conductivities, they typically exhibit low Seebeck coefficients. In this work, we create a database and apply machine learning techniques to identify metallic binary alloys with large Seebeck coefficients. Specifically, we identify Ni-Fe as a promising candidate for active cooling. We then fabricate Ni-Fe ingots and demonstrate thermoelectric power factor values as high as 120 μW/cm·K2 at 200 K for these stable alloys, which are composed of cost-effective and abundant elements. Furthermore, we show that the effective thermal conductivity of these alloys, under small temperature differences, can exceed that of pure copper at temperatures above 250 K. Copyright © 2025, The Authors. All rights reserved.

关键词： Iron alloys

来源：评论

学校读者我要写书评

暂无评论

Crucian carp position and pose recognition based on an improved photo-model method 3

Crucian carp position and pose recognition based on an impro...

引用

2022 3rd International Conference on Signal Processing and computer science, SPCS 2022

作者： Wen, Changyi Wang, Jirong Dai, Xiangwu Li, Jun College of Mechanical and Electrical Engineering Qingdao University Qingdao266071 China College of Computer Science and Technology Qingdao University Qingdao266071 China Weihai Innovation Research Institute of Qingdao University Weihai264200 China

This research aimed to solve the problem of fish position and pose recognition in the process of robot grasping fish. Crucian carp is the subject of this research and studies the position and pose recognition algorithm of crucian carp by photo-model-based stereo-vision. This research aims at the problem that the fitness algorithm has a poor effect in identifying photo models with more black and white information, this research proposes an improved fitness algorithm. The new fitness algorithm uses the improved HSV model and adds the Saturation and Value to the evaluation value of the original fitness algorithm. Then Real-Time Multi-step Genetic Algorithm (RM-GA) is used to verify the improvement effect of the fitness algorithm. We use the improved algorithm for experiments. The findings indicated that the peak fitness is significantly improved, with an average increase of about 95%. At the same time, the position and posture recognition error of crucian carp is reduced by more than 50%, and the real-time performance of the algorithm is improved. © Published under licence by IOP Publishing Ltd.

关键词： Fish

来源：评论

学校读者我要写书评

暂无评论

On UAV-IRS-Aided Covert Communication of D2D-Enabled Cellular Networks

On UAV-IRS-Aided Covert Communication of D2D-Enabled Cellula...

引用

International Conference on Networking and Network Applications (NaNA)

作者： Sibo Wang Bin Yang Wu Wang Shikai Shen Xiaohong Jiang School of Mechanical Engineering Jiangsu University Zhenjiang China School of Computer and Information Engineering Chuzhou University Chuzhou China School of Mathematics and Computer Science Yunnan Minzu University Kunming China School of Information Engineering Kunming University Kunming China School of Systems Information Science Future University Hakodate Japan

In this paper, we investigated covert communication in D2D-Enabled cellular network scenario and proposed to improve the covert transmission rate with the assistance of a UAV-IRS (Unmanned Aerial Vehicle with Intelligent Reflecting Surface) mobile relay. The use of UAV-IRS enables D2D transmitters on the ground to transmit sensitive information covertly to D2D receivers without detection by warden Willie. We first determine an optimal detection threshold and derive the detection error probability at Willie, this probability represents the worst-case legitimate transmission. Then, we maximize the covert transmission rate through the maximum transmission power of D2D transmitters, optimal phase shift and optimal horizontal position of UAV-IRS. Simulation results are presented the effectiveness of UAV-IRS-aided covert communication of D2D-enabled cellular networks.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3D Class A Bézier curves with monotone curvature

引用

CAD computer Aided Design 2023年 159卷

作者： Wang, Aizeng He, Chuan Zheng, Jianmin Zhao, Gang School of Mechanical Engineering & Automation Beihang University Beijing100191 China State Key Laboratory of Virtual Reality Technology & Systems Beihang University Beijing100191 China School of Computer Science and Engineering Nanyang Technological University Singapore

This paper introduces a special class of 3D Bézier curves that are defined by their degree, a starting point, the first leg of their control polygons, and a 3D affine transformation composing of a uniform scaling and a rotation. We present new formulas for the curvature of such Bézier curves and based on the new formulas we derive sufficient conditions for the curves to have monotonic curvature. The conditions are expressed by a simple constraint on the rotation angle and the scaling factor. This facilitates constructing 3D Class A Bézier curves that are a generalization of planar typical curves proposed by Mineur et al. (1998), which are often used in automotive and other design applications. Some examples are provided to demonstrate the effectiveness of our construction. © 2023 Elsevier Ltd

关键词： Geometry

来源：评论

学校读者我要写书评

暂无评论

Vision Transformer Based Automated Model for Enhancing Lung Cancer Classification

Vision Transformer Based Automated Model for Enhancing Lung ...

引用

IEEE International Workshop on Imaging Systems and Techniques (IST)

作者： Akbar Sheikh Akbari Arvind Kumar B Ramachandra Reddy Koushlendra Kumar Singh Masahiro Takei Leeds Beckett University UK Machine Vision and Intelligence Lab National Institute of Technology Jamshedpur Jharkhand India Computer Science and Engineering National Institute of Technology Jamshedpur Jharkhand India Department of Mechanical Engineering Chiba University Chiba Japan

ISBN: (数字)9798350378214

ISBN: (纸本)9798350378221

Lung cancer is one of the leading causes of cancer related mortality. The early detection and classification of the cancers tissues will reduce the mortalities rate. The present research focus on the development of automated classification model for lung and colon cancers tissues based on the histopathology images. The present work encompasses a vision transformer (ViT) based model to enhance diagnostic accuracy of lung cancers tissues. The proposed model utilizes the self-attention mechanism of ViT to focus on essential features present in histopathologicals images. The proposed model has been validated using two different dataset namely LC25000 & IQ-OTH/NCCD with 25000 & 1096 images respectively. The performance of proposed model is compared with traditional convolutional neural network (CNN) model and it has been observed the based model outforms better in terms of accuracy which - 98.80% & 99.09% respectively for datasets.

关键词： computer vision Analytical models Accuracy Lungs Lung cancer Imaging Transformers Convolutional neural networks Reliability Testing

来源：评论

学校读者我要写书评

暂无评论

A Lecture Centric Automated Distractor Generation for Post-Graduate Software engineering Courses

A Lecture Centric Automated Distractor Generation for Post-G...

引用

Frontiers of Information Technology (FIT)

作者： Rudeema Chughtai Farooque Azam Muhammad Waseem Anwar Wasi Haider But Muhammad Umar Farooq Department of Computer & Software Engineering College of Electrical & Mechanical Engineering (CEME) National University of Science & Technology (NUST) Islamabad Pakistan

ISBN: (纸本)9798350345940

Critical circumstances, natural disasters or pandemics like COVID 19 gave rise to the wide applicability of E-learning into education system. Efficient and fair online assessment is very important to utilize the inevitable benefits of E-learning.. In order to make it efficient, the trend of assessment has shifted from the subjective type to the objective type assessments which is mainly based on Multiple Choice Questions (MCQs), generation of MCQs is a tedious, tiresome and time consuming task. To cater this dire need, this study proposes an automated Multiple Choice Question (MCQ) generation by utilizing state of the art transformer based model T5 for the task of question generation and a lexicon based approach Sense2vec for the task of distractor generation. It also presented a domain specific lecture text based test data for performing evaluation on the task of domain specific lecture text based MCQ generation.

关键词： COVID-19 Electronic learning Pandemics Education Transformers Market research Task analysis

来源：评论

学校读者我要写书评

暂无评论

Dynamic Incentive Design for Demand Response in Energy System MicroGrids 25

Dynamic Incentive Design for Demand Response in Energy Syste...

引用

Proceedings of the 16th ACM International Conference on Future and Sustainable Energy Systems

作者： Millend Roy Vladimir Pyltsov Ruturaj Godse Tony Dear Operations Research Columbia University New York NY USA Mechanical Engineering Columbia University New York NY USA Computer Science Columbia University New York NY USA

来源：评论

学校读者我要写书评

暂无评论

A Novel Scatter-enhanced Correlation Feature Learning Method

A Novel Scatter-enhanced Correlation Feature Learning Method

引用

2021 IEEE International Conference on Artificial Intelligence and computer Applications, ICAICA 2021

作者： Su, Shuzhi Xie, Jun Zhu, Yanmin Liang, Xingzhu Anhui University of Science Technology School of Computer Science and Engineering Huainan China Institute of Energy Hefei Comprehensive National Science Center Hefei China Anhui University of Science Technology School of Mechanical Engineering Huainan China Wuhu China

ISBN: (纸本)9781665418676

Canonical Correlation Analysis (CCA) is an essential algorithm in the feature learning field. However, it does not utilize supervised information, and it failed to solve nonlinear problems. Therefore, this paper proposes a novel feature learning algorithm called Scatter-enhanced Canonical Correlation Analysis (SeCCA). This paper integrates the internal structure information and supervised information of the data and embeds them into the canonical correlation framework. The excellent image recognition performance of this algorithm can be demonstrated by extensive experimental results. © 2021 IEEE.

关键词： Learning algorithms

来源：评论

学校读者我要写书评

暂无评论

Applying different machine learning algorithms to predict the viscosity behavior of MWCNT–alumina/water–ethylene glycol (80:20) hybrid antifreeze

引用

International Journal of Thermofluids 2024年 24卷

作者： Hussein, Shaymaa Abed Omar, Ihab Saddam, Ali B. Baghoolizadeh, Mohammadreza Salahshour, Soheil Pirmoradian, Mostafa Department of pathological analyzes Al Manara College for Medical Sciences Maysan Iraq Air Conditioning Engineering Department Faculty of Engineering Warith Al-Anbiyaa University Karbala56001 Iraq Department of Electrical Engineering Techniques Al-Amarah University College Maysan Iraq Department of Mechanical Engineering Shahrekord University Shahrekord88186-34141 Iran Faculty of Engineering and Natural Sciences Istanbul Okan University Istanbul Turkey Faculty of Engineering and Natural Sciences Bahcesehir University Istanbul Turkey Department of Computer Science and Mathematics Lebanese American University Beirut Lebanon Department of Mechanical Engineering Khomeinishahr branch Islamic Azad University Khomeinishahr Iran

While machine learning has become the new way of analyzing data, neutral networks form the basis of this revolutionary technology. In this work, we shall employ the power of neural networks to analyze and demystify the processes in nanofluids. By combining the precision of neural networks with the optimization capabilities of genetic algorithms, we aim to create a more accurate and efficient prediction model for MWCNT-alumina/water-ethylene glycol (80:20) hybrid antifreeze. Our approach entails using an MLP neural network and several training functions (LM, GD, BFGS, BN) with an adjustable number of neurons. The inputs of the network are φ (solid volume fraction or ), temperature (T), and shear rate (γ), and the output is μnf of MWCNT-alumina/water-ethylene glycol (80:20) hybrid anti-freeze. To improve the accuracy of the final model, we use genetic optimization to make final adjustments to the parameters of the neural network. Utilizing the detailed analysis of the primary characteristics of these algorithms, we conclude that the BFGS function is the best to obtain neural network training. Steady performance achieved by this function—0.99828 of the R-value and RMSE value significantly equal to 0.213—illustrates good stability and accuracy of the suggested model. This work contributes to progressing the existing knowledge about the behavior of nanofluids and can stimulate further improvement in heat transfer and energy utilization. © 2024 The Author(s)

关键词： Genetic 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：