检索结果-内蒙古大学图书馆

8th International Conference on Intelligent Informatics and Biomedical Sciences, ICIIBMS 2023

作者： Cao, Guogang Peng, Zeyu Wu, Yan Zhou, Zhilin Shanghai Institute of Technology SIT Shanghai China Shanghai Institute of Technology School of Computer Science and Information Engineering Shanghai China

ISBN: (纸本)9798350369564

Retinal OCT image segmentation plays a crucial role in the diagnosis of eye diseases. Traditional layer segmentation methods are time-consuming and heavily reliant on the subjective judgment of annotators. To address these challenges, this paper proposes an automatic retinal OCT image layer segmentation approach based on attention mechanisms. The model replaces regular convolutions of U-Net with dilated convolutions to increase the model's receptive field while preserving the feature map size. Additionally, a hybrid attention mechanism module, called spatial and channel attention module (scSE), is introduced in encoder part scSE module combines spatial and channel attention, enhancing the model's focus on important features and improving its robustness and generalization. The proposed model is validated on a publicly available dataset, and the results demonstrate its superiority over other methods in retinal layer segmentation, achieving a Dice coefficient of 0.845. The proposed model performs well in retinal layer segmentation tasks, providing more accurate guidance for assisting doctors in diagnosing retinal diseases. © 2023 IEEE.

关键词： Convolution

来源：评论

学校读者我要写书评

暂无评论

A Barrier-Based Machine Learning Approach for Intrusion Detection in Wireless Sensor Networks

引用

computers, Materials & Continua 2025年第3期82卷 4181-4218页

作者： Haydar Abdulameer Marhoon Rafid Sagban Atheer Y.Oudah Saadaldeen Rashid Ahmed Information and Communication Technology Research Group Scientific Research CenterAl-Ayen UniversityThi-Qar64011Iraq College of Computer Sciences and Information Technology University of KerbalaKarbala56001Iraq Enginerring Technical College Al-Ayen UniversityThi-Qar64011Iraq Information Technology College University of BabylonHilla51002Iraq Department of Computer Science College of Education for Pure ScienceUniversity ofThi-QarNasiriyah64001Iraq Artificial Intelligence Engineering Department College of EngineeringAl-Ayen UniversityThi-Qar64001Iraq Department of Computer Science Bayan UniversityErbilKurdistan44001Iraq

In order to address the critical security challenges inherent to Wireless Sensor Networks(WSNs),this paper presents a groundbreaking barrier-based machine learning *** applications like military operations,healthcare monitoring,and environmental surveillance increasingly deploy WSNs,recognizing the critical importance of effective intrusion detection in protecting sensitive data and maintaining operational *** proposed method innovatively partitions the network into logical segments or virtual barriers,allowing for targeted monitoring and data collection that aligns with specific traffic *** approach not only improves the *** are more types of data in the training set,and this method uses more advanced machine learning models,like Convolutional Neural Networks(CNNs)and Long Short-Term Memory(LSTM)networks together,to see coIn our work,we used five different types of machine learning *** are the forward artificial neural network(ANN),the CNN-LSTM hybrid models,the LR meta-model for linear regression,the Extreme Gradient Boosting(XGB)regression,and the ensemble *** implemented Random Forest(RF),Gradient Boosting,and XGBoost as baseline *** train and evaluate the five models,we used four possible features:the size of the circular area,the sensing range,the communication range,and the number of sensors for both Gaussian and uniform sensor *** used Monte Carlo simulations to extract these *** on the comparison,the CNN-LSTM model with Gaussian distribution performs best,with an R-squared value of 99%and Root mean square error(RMSE)of 6.36%,outperforming all the other models.

关键词： Intrusion detection system(IDS) hybrid models of CNN-LSTM WSN extreme gradient boosting(XGBoost)regressor ensemble model

来源：评论

学校读者我要写书评

暂无评论

Building Crack Detection Using CNN and Class Activation Mapping

Building Crack Detection Using CNN and Class Activation Mapp...

引用

2024 International Conference on Artificial Intelligence and Quantum Computation-Based Sensor Applications, ICAIQSA 2024

作者： Keerthana, Bodduru Devi, Jathuka Aruna Madhuri, Surisetty Rao, Marada Srinivasa Rao, Koduganti Venkata Geethika, P.S.S. Anil Neerukonda Institute of Technology and Sciences Department of Information Technology Andhra Prades Visakhapatnam India Andhra Prades Visakhapatnam India Vignan Institite of Engineering for Women Department of Information Technology Andhra Prades Visakhapatnam India MVGR College of Engineering Department of Information Technology Andhra Prades Vizianagaram India Guru Nanak Institutions Technical Campus Department of Computer Science and Engineering Hyderabad India

ISBN: (纸本)9798331517953

As the population increases, further structures are being erected to meet the demands of the populace. Nonetheless, this surge in development presents the challenge of guaranteeing that each edifice fulfills the aspirations of its inhabitants. Identifying and rectifying damage or defects in structures can be challenging, particularly when conventional techniques typically necessitate comprehensive examinations to pinpoint issues. To address this difficulty, we are utilizing advanced technology such as Deep Learning models, particularly Convolutional Neural Networks (CNNs), along with approaches like Class Activation Mapping (CAM). These models utilize extensive datasets to learn and recognize patterns, enabling them to precisely identify specific forms of damage or defects in structures. We intend to forecast probable damage rates in advance through the application of these Deep Learning approaches. This proactive strategy enables us to avert catastrophic incidents by recognizing and resolving problems prior to their escalation. The great precision of these models ensures that buildings are safe and fulfill tenant expectations, hence enhancing the safety and reliability of structures within our communities. Our model attained an accuracy of 88%. © 2024 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Recognition of Wh-Question Sign Gestures in Video Streams using an Attention Driven C3D-BiLSTM Network 2

Recognition of Wh-Question Sign Gestures in Video Streams us...

引用

2nd International Conference on Machine Learning and Data engineering, ICMLDE 2023

作者： Dey, Arnab Biswas, Samit Le, Dac-Nhuong Dept. of Computer Science and Technology Indian Institute of Engineering Science and Technology Howrah Shibpur711103 India Faculty of Information Technology Haiphong University Haiphong180000 Viet Nam

Human sign language is a visual and gestural means of communication used by people with hearing impairments to interact with others. It has the potential to enable interaction between individuals who struggle with verbal communication and those who lack the essential proficiency to comprehend sign language. Wh-questions constitute a substantial portion of daily sign language interactions. Automatic recognition of Wh-question words in sign language, encompassing interrogative words, has received limited attention in gesture recognition. In this context, a novel dataset named the American Question Sign Video dataset (AQSVd) has been introduced as a significant contribution. Recognizing Wh-question signs in video streams empowers individuals to effectively convey messages through hand movements and gestures, fostering inclusivity and accessibility in communication for deaf and hearing-impaired individuals. The paper proposes a Deep Convolutional-3D BiLSTM Multi-head Attention network for recognizing American Wh-Question word sign gestures in video streams. Incorporating Multi-head attention in the proposed model enhances its ability to capture intricate spatial and temporal features, essential for accurate gesture recognition. The proposed model is thoroughly evaluated across a range of datasets, such as INCLUDE50, WLASL-100, and our AQSVd dataset, achieving remarkable results. Specifically, our model demonstrated an outstanding 98.91% validation accuracy on the AQSVd dataset. The C3D-BiLSTM MHAttention model outperforms other state-of-the-art models, demonstrating its superiority in sign language recognition tasks. The proposed dataset and C3D-BiLSTM Multi-head Attention model contribute significantly to this field, offering potential benefits in education, human-robot interaction, and overall communication for the hearing-impaired community. © 2024 Elsevier B.V.. All rights reserved.

关键词： Gesture recognition

来源：评论

学校读者我要写书评

暂无评论

On the Performance of Multidimensional Constellation Shaping for Linear and Nonlinear Optical Fiber Channel 49

On the Performance of Multidimensional Constellation Shaping...

引用

49th European Conference on Optical Communications, ECOC 2023

作者： Chen, Bin Liang, Zhiwei Li, Shen Lei, Yi Liga, Gabriele Alvarado, Alex School of Computer Science and Information Engineering Hefei University of Technology China Department of Electrical Engineering Chalmers University of Technology Gothenburg Sweden Department of Electrical Engineering Eindhoven University of Technology Netherlands

ISBN: (纸本)9781839539954

Multidimensional constellation shaping of up to 32 dimensions with different spectral efficiencies are compared through AWGN and fiber-optic s imulations. The results show that no constellation is universal and the balance of required and effective SNRs should be jointly considered for the specific optical transmission scenario. © The Institution of engineering & technology 2023.

关键词： Optical fibers

来源：评论

学校读者我要写书评

暂无评论

EPCA—Enhanced Principal Component Analysis for Medical Data Dimensionality Reduction

引用

SN computer Science 2023年第3期4卷 1-14页

作者： Vinutha, M.R. Chandrika, J. Krishnan, Balachandran Kokatnoor, Sujatha Arun Department of Information Science and Engineering Malnad College of Engineering Hassan 573202 India Department of Computer Science and Engineering School of Engineering and Technology CHRIST (Deemed to be University) Bangalore 560074 India

Innovations in technology from the last one decade have led to the generation of colossal amounts of medical data with comparably low cost. Medical data should be collected with utmost care. Sometimes, the data have high features but not all the features play an important role in drawing the relations to the mining task. For the training of machine learning algorithms, all the attributes in the data set are not relevant. Some of the characteristics may be negligible and some characteristics may not influence the outcome of the forecast. The pressure on machine learning algorithms can be minimized by ignoring or taking out the irrelevant attributes. Reducing the attributes must be done at the risk of information loss. In this research work, an Enhanced Principal Component Analysis (EPCA) is proposed, which reduces the dimensions of the medical dataset and takes paramount care of not losing important information, thereby achieving good and enhanced outcomes. The prominent dimensionality reduction techniques such as Principal Component Analysis (PCA), Singular Value Decomposition (SVD), Partial Least Squares (PLS), Random Forest, Logistic Regression, Decision Tree and the proposed EPCA are investigated on the following Machine Learning (ML) algorithms: Support Vector Machine (SVM), Artificial Neural Networks (ANN), Naïve Bayes (NB) and Ensemble ANN (EANN) using statistical metrics such as F1 score, precision, accuracy and recall. To optimize the distribution of the data in the low-dimensional representation, EPCA directly mapped the data to a space with fewer dimensions. This is a result of feature correlation, which made it easier to recognize patterns. Additionally, because the dataset under consideration was multicollinear, EPCA aided in speeding computation by lowering the data's dimensionality and thereby enhanced the classification model's accuracy. Due to these reasons, the experimental results showed that the proposed EPCA dimensionality reduction technique per

关键词： Accuracy Artificial neural networks Liver cirrhosis Machine learning Medical data Principal component analysis Singular value decomposition Support vector machine

来源：评论

学校读者我要写书评

暂无评论

Causal Policy Gradient for End-to-End Communication Systems 16

Causal Policy Gradient for End-to-End Communication Systems

引用

16th International Conference on COMmunication Systems and NETworkS, COMSNETS 2024

作者： Shirodkar, Shounak Banerjee, Serene International Institute of Information Technology Computer Science and Engineering Bangalore India Ericsson Research Machine Reasoning and Hybrid Ai Bangalore India

ISBN: (纸本)9798350383119

End-to-End (E2E) learning of communication systems can be realized with autoencoders when the channel model is differentiable. However, due to various time-varying unknown factors such as signal attenuation, hardware distortions, and multi-path fading, the channel model in practice is often non-differentiable. For such cases, the state-of-the-art demonstrates that it is possible to learn E2E reconstruction of the transmitted data, using Gaussian policy gradient or Deep Deterministic Policy Gradient (DDPG) algorithms. However, a few shortcomings of the proposed approaches include: (a) the solution not scaling well with the number of channel uses, (b) needing long coherence time of the channel compared to the rate at which examples can be processed at training, (c) poor sample efficiency during training and slow convergence especially for larger constellations, and (d) poor generalization. Uncovering the causal relationships between: the state (i.e. transmitted symbols), the action (i.e. constellation shaping), and the reward (i.e. the loss between the received and the transmitted symbols), has potential to address the above limitations. We propose two algorithms for causality-inspired end-to-end learning, where the causal relationship between: (a) the state and the action, and (b) the action and the reward shows up to 90% faster convergence, compared to the state-of-the-art variants of policy gradient algorithms. For the trained model, the Block Error Rate (BLER) at 17dB Signal to Noise Ratio (SNR) is less than 0.005, for various constellation shapes. To the best of our knowledge, this is the first demonstration of causality-inspired end-to-end learning of communication systems, without a prior channel model. © 2024 IEEE.

关键词： causal reinforcement learning domain generalization End-to-end learning of communication systems non-differentiable channel model sample-efficient learning

来源：评论

学校读者我要写书评

暂无评论

Recommender Process Based on Trust-Distrust Factor for Signed Social Networks 12th

Recommender Process Based on Trust-Distrust Factor for Signe...

引用

12th International Conference on Soft Computing for Problem Solving, SocProS 2023

作者： Lal, Roshan Sharma, Sanjay Kumar Department of Computer Science and Engineering University School of Information and Communication Technology Gautam Buddha University Uttar Pradesh Greater Noida India

ISBN: (纸本)9789819732913

Recommender process have gained popularity in recent years due to their effectiveness in addressing the issue of information overload. It is nonetheless susceptible to several intrinsic problems, such as cold starting and data scarcity. Numerous studies show that leveraging data from social networks is a highly successful tactic to deal with such issues. The interaction between users based on their conduct and social links is also taken into consideration in research on the recommendation process that integrates social interactions, in addition to the user's preferences for the product. Social network relationships that involve both trust and distrust have not grabbed much attention. Our research finding suggests the methods of integrating the trust and distrust social relationships using machine learning in order to enhance the collaborative filtering recommendation algorithm, which combined the users’ trust and distrust social relationships and effectively reduced the sparseness in the signed social networks. The experimental results shows that the recommended methodologies outperform state-of-the-art algorithms based on trust-distrust using machine learning. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2024.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Bibliometric Analysis of research trends on Graph Neural Networks 20

Bibliometric Analysis of research trends on Graph Neural Net...

引用

20th CSI International Symposium on Artificial Intelligence and Signal Processing, AISP 2024

作者： Alavi, Meysam Valiollahi, Arefeh Pakravan, Mansooreh Tarbiat Modares University Information Technology Dept Tehran Iran Tarbiat Modares University Electrical and Computer Engineering Dep Tehran Iran

ISBN: (纸本)9798350383942

Graph Neural Networks (GNNs) are a powerful tool for analyzing complex systems and irregular data structures like graphs, revolutionizing tasks like node classification and link prediction. Therefore, regularly assessing scientific outputs to understand the current landscape is crucial. In this regard, scientometric analysis is utilized to assess, interpret, and anticipate the scientific standing of researchers, research institutions, academic journals, and countries on a global scale. Utilizing bibliometric techniques like distribution analysis, co-word networks, and scientific cooperations has proven valuable in unveiling the production trends of researchers in a scientific field, revealing both explicit and concealed dimensions. Despite existing surveys, a significant gap exists in the literature regarding bibliometric studies on the Web of Science (WoS) database in the emerging field of GNNs. This paper uses bibliometric analysis on WoS to explore the evolving GNN research area, covering contributors, collaborations, top journals, and trends identified through keywords and high citations. The research reveals that China, the United States, and Australia lead in publications on neural graph networks. Additionally, the findings underscore a predominant focus on graph neural networks within the realm of computer science. © 2024 IEEE.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Book Genre Classification System Using Machine Learning Approach: A Survey 5th

Book Genre Classification System Using Machine Learning Appr...

引用

5th International Conference on Soft Computing and Signal Processing, ICSCSP 2022

作者： Sethy, Abhisek Rout, Ajit Kumar Gunda, Chandra Routhu, Sai Kiran Kallepalli, Srinu Garbhapu, Geetha Department of Information Technology GMR Institute of Technology Srikakulam Rajam India Department of Computer Science & Engineering Silicon Institute of Technology Bhubaneswar India

ISBN: (纸本)9789811986680

The amount of complicated texts and documents that require a deeper understanding of machine learning techniques has expanded rapidly in recent decades. Several machine learning approaches have shown exceptional results in NLP. Complex models and non-linear data correlations help these learning systems function well. It is difficult for academics to discover appropriate text classification structures, designs, and procedures. Manually reading and classifying a book's category would be tedious. It takes a long time for a language beginner to absorb the complete text in order to discern its genre. Natural Language Processing (NLP), which is commonly used nowadays, can help overcome this issue. It includes text classification and summarization. The book synopsis is input to the Machine Learning algorithms, which output the genre. This review includes text feature extraction, dimensionality reduction, algorithms, and evaluation. Finally, the limitations of each strategy are assessed, as well as their practicality. © 2023, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Support vector machines

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：