检索结果-内蒙古大学图书馆

Innovative Approaches in Smart Technologies (ISAS), International Symposium on

作者： Elsadig Ahmed Mohamed Babiker Hanan Hassan Ali Adlan Dept. of Computer Science College of Computer Engineering and Science Prince Sattam Bin Abdulaziz University Kharj KSA Dept. of Computer Science Faculty of Mathematical Sciences Uof K Dept. of Computer Science College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University Riyadh KSA

ISBN: (数字)9798331540104

ISBN: (纸本)9798331540111

Neural Networks are the state-of-the-art models that derive intelligent systems. Generalization of the neural network model is a judge of how well an architecture mimic human intelligence. Self-extraction neural networks become the dominant neural networks, especially in autonomous systems such as intelligent robotics, autonomous vehicles, ⃛ etc. This paper presents a neural network based on distance measure, ElHaNet Neural Network. ElHaNet network is a self-extraction neural network and is capable of pattern recognition. ElHaNet architecture composed of an extraction phase and a classification phase. Neurons in the extraction network capture patterns in the presented input. The classification network performs on the input patterns to be classified according to the problem in hand. The developed architecture is bench marked against well-known architectures for hand written digit recognition. The performance of the network is found to compete favorably with state-of-the-art models in the literature.

关键词： Hands Human intelligence Neurons Pattern recognition Biological neural networks Robots Autonomous vehicles

来源：评论

学校读者我要写书评

暂无评论

Utilizing low-cost optical sensor for the measurement of particulate matter and calculating Pearson's correlation coefficient 16

Utilizing low-cost optical sensor for the measurement of par...

引用

16th International Conference on Correlation Optics, COR 2023

作者： Kirešová, Simona Rusyn, Volodymyr Guzan, Milan Vorobets, George Sobota, Branislav Vorobets, Oleksandr Dept. of Theoretical and Industrial Electrical Engineering Technical University of Košice Letná 9 Košice042 00 Slovakia Dept. of Radio Engineering and Information Security Yuriy Fedkovych Chernivtsi National University 2 Kotsjubynskyi Str. Chernivtsi58000 Ukraine Dept. of Computer Systems and Networks Yuriy Fedkovych Chernivtsi National University 2 Kotsjubynskyi Str. Chernivtsi58000 Ukraine Dept. of Computers and Informatics Technical University of Košice Letná 9 Košice042 00 Slovakia

ISBN: (纸本)9781510671829

This article discusses the possibilities of measuring Particulate Matter using optical low-cost sensors. Depending on the sensor used, not only can there be problems with the positioning of the sensor with respect to the incoming sunlight, but it may also be possible to measure ultrafine particles. While PM2.5 and PM10 are commonly used to characterize air quality, in this paper we point out the need to pay more attention to smaller particles up to 0.1 μm in size (also referred to as ultrafine particulates) due to their highly negative effects on the human health. Attention is also paid to the correlation between particulate matter and meteorological factors like pressure, temperature, and humidity. However, the sudden change in Pearson's correlation coefficient unveiled a need to look into the effect of the wind on particulate matter. © COPYRIGHT SPIE. Downloading of the abstract is permitted for personal use only.

关键词： Optical sensors

来源：评论

学校读者我要写书评

暂无评论

A Digital Healthcare Service Architecture for Seniors Safety Monitoring in Metaverse

A Digital Healthcare Service Architecture for Seniors Safety...

引用

2023 IEEE International Conference on Metaverse Computing, Networking and Applications, MetaCom 2023

作者： Qu, Qian Xu, Ronghua Sun, Han Chen, Yu Sarkar, Sumantra Ray, Indrajit Binghamton University Dept. of Electrical and Computer Engineering BinghamtonNY13902 United States Binghamton University School of Management BinghamtonNY13902 United States Colorado State University Dept. of Computer Science Fort CollinsCO80523 United States

ISBN: (纸本)9798350333336

We have been witnessing an unprecedented increase in the aging population in human history. It is nontrivial to ensure the health and safety of seniors living alone. The prohibitive human labor cost necessitates more sustainable, technology-oriented approaches instead of labor-intensive solutions. The raising digital healthcare services (DHS) leveraging the Internet of Medical Things (IoMT), Digital Twins (DT), and advanced fifth-generation and beyond (B5G) wireless communication technology, are widely recognized as promising solutions. By enabling a seamless interwoven of the physical world and cyberspace, Metaverse makes an ideal home for the next generation of DHS. Thanks to characteristics of decentralization, traceability, and unalterability, Blockchain is envisioned to enhance security properties in Metaverse. This paper proposes MetaSafe, a DHS architecture for seniors' safety monitoring in Metaverse. Based on monitoring data collected by sensors, the activities and status of seniors, who are considered as the physical objects (PO), are mirrored to corresponding logical objects (LO) in a virtual community in the Metaverse, where activity recognition, potential risk prediction, and alert generation are realized. By leveraging Non-Fungible Token (NFT) technology to tokenize identities (POs and LOs) and data streams of the DHS on the blockchain, an NFT-based authentication fabric allows for verifiable ownership and traceable transferability during the data-sharing process. Specifically, an instant alerting system is introduced in this work that leverages a hybrid algorithm combining the singular spectrum analysis (SSA) approach with the long-short-term memory (LSTM) networks. Through an extensive experimental study, MetaSafe is validated as a feasible and promising approach to protect seniors living alone. © 2023 IEEE.

关键词： Blockchain

来源：评论

学校读者我要写书评

暂无评论

A Lightweight CNN-SVM Explainable AI Approach for Classification and Visualization of Grape Leaf Disease

A Lightweight CNN-SVM Explainable AI Approach for Classifica...

引用

International Conference on Advancement in Electrical and Electronic engineering (ICAEEE)

作者： Hasibul Islam Peyal Zannatul Mawa Leion Md. Nazmul Abdal Md. Injamul Islam Shafiun Miraz Mahmudur Rashid Remon Md. Mahin Rahaman Kontho Nishat Tasnim Dept. of Electrical & Computer Engineering Rajshahi University of Engineering & Technology Rajshahi Bangladesh Dept. of Computer Science and Engineering University of Liberal Arts Bangladesh Dhaka Bangladesh Dept. of Computer Science & Engineering Uttara University Dhaka Bangladesh Rajshahi University of Engineering & Technology Rajshahi Bangladesh Dept. of Electrical and Electronic Engineering North South University Dhaka Bangladesh Dept. of Agroforestry and Environmental Science Sher-e-Bangla Agricultural University Dhaka Bangladesh

ISBN: (数字)9798350388282

ISBN: (纸本)9798350388299

Grape is highly esteemed as a significant agricultural crop in Bangladesh. Plant diseases primarily result from the presence of pathogens and pest insects, leading to a significant decline in productivity if not properly addressed. The fundamental aim of this research is to employ a streamlined CNN-SVM architecture, utilizing deep learning techniques, to accurately categorize grape leaves into three distinct disease classes and one healthy class. The proposed model surpasses the accuracy of the previously trained transfer learning models VGG-16 and VGG-19 while having approximately $257\times$ to $267\times$ times fewer parameters (0.537 M). On average, the proposed model achieves a classification accuracy of 99.18%, which is significantly higher than the 93.42% and 91.94% achieved by the transfer learning models, respectively. With a precision, recall, and F1 score close to 99%, the suggested model provides excellent results. The model’s outstanding performance is further validated by its remarkable Area Under Curve (AUC) score of 99.98%. In addition to using less disc space (about 6 MB), the suggested model because of being lightweight shows a significant decrease in parameters. To visually display the disease identified by the proposed model, a transparent AI methodology has been utilized, specifically the Gradient Weighted Class Activation Mapping (Grad-CAM) technique. To better understand which area was responsible for the classification, a heatmap has been created.

关键词： Productivity Heating systems Visualization Plant diseases Pathogens Accuracy Computational modeling

来源：评论

学校读者我要写书评

暂无评论

BERT-D2: Drug-Drug Interaction Extraction using BERT

BERT-D2: Drug-Drug Interaction Extraction using BERT

引用

2022 International Conference for Advancement in Technology, ICONAT 2022

作者： Datta, Tanmoy Tapos Shill, Pintu Chandra Al Nazi, Zabir Khulna University of Engineering Technology Dept. of Computer Science and Engineering Khulna Bangladesh Brainekt AI Lab Dhaka Bangladesh

ISBN: (纸本)9781665425773

When one drug interacts with another, it is known as a drug-drug interaction. This could change how one or both drugs work in the body, or induce unforeseen adverse effects. The performance of a combination can be effectively degraded or improved by mixing different medications. In some cases, it can adversely affect the patient's health. So, it's of great importance and time demanding to classify the interaction of drugs. Drug-drug interaction is amongst the top challenging and far-reaching applications of natural language processing. In this work, we have presented drug-drug interaction using the BERT (Bidirectional Encoder Representations from Transformers) model. The previous state of the art performance on the DDI Extraction 2013 corpus was using different variations of convolutional neural networks and LSTMs. In order to validate our proposed model, well-known benchmark data set are used and our BERT-based classification has achieved a much higher score than previous methods at 90.69% accuracy and 81.97% f1-score. © 2022 IEEE.

关键词： Extraction

来源：评论

学校读者我要写书评

暂无评论

Multi-Point Positional Insertion Tuning for Small Object Detection

arXiv

引用

arXiv 2024年

作者： Goto, Kanoko Karasawa, Takumi Hirose, Takumi Kawakami, Rei Inoue, Nakamasa Dept. of Computer Science Institute of Science Tokyo Tokyo Japan Dept. of Systems and Control Engineering Institute of Science Tokyo Tokyo Japan

Small object detection aims to localize and classify small objects within images. With recent advances in large-scale vision-language pretraining, finetuning pretrained object detection models has emerged as a promising approach. However, finetuning large models is computationally and memory expensive. To address this issue, this paper introduces multi-point positional insertion (MPI) tuning, a parameter-efficient finetuning (PEFT) method for small object detection. Specifically, MPI incorporates multiple positional embeddings into a frozen pretrained model, enabling the efficient detection of small objects by providing precise positional information to latent features. Through experiments, we demonstrated the effectiveness of the proposed method on the SODA-D dataset. MPI performed comparably to conventional PEFT methods, including CoOp and VPT, while significantly reducing the number of parameters that need to be tuned. Copyright © 2024, The Authors. All rights reserved.

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Detection of Brain Tumor Using a Deep Learning Approach

Detection of Brain Tumor Using a Deep Learning Approach

引用

Artificial Intelligence and Data engineering (AIDE), International Conference on

作者： Arun Kumar S Anand Kumar M Robin Rohit Vincent Anusha M S Uttam Upendra Hegde N Nauman Pasha Dept. of Computer Studies Symbiosis Institute of Computer Studies and Research (SICSR) Pune India Dept. of Information Science and Engineering Networks & IoT Lab Presidency University Bangaluru India Dept. of Computer Science and Engineering Bio-Intelligence Lab Presidency University Bangaluru India

ISBN: (数字)9798331527518

ISBN: (纸本)9798331527525

This study introduces a deep learning-based method for classifying brain tumors using a pre-trained VGG19 convolutional neural network (CNN). By leveraging transfer learning, we adapted the VGG19 model with custom fully connected layers and trained it on MRI images to differentiate between tumor and non-tumor cases. The model’s architecture includes dropout layers to minimize overfitting and enhance generalization while maintaining the robust feature extraction capabilities of the VGG19 base. We developed a Flask web application interface to make the model accessible via web-based user input, allowing users to upload MRI images and receive immediate predictions on tumor presence. Preliminary results show that proposed model can accurately classify brain tumor images with high accuracy, highlighting the potential of deep learning in automating brain tumor diagnostics. This tool could significantly aid clinicians in making rapid and accurate diagnoses, ultimately improving patient outcomes.

关键词： Deep learning Adaptation models Accuracy Magnetic resonance imaging Transfer learning Brain tumors Predictive models Brain modeling Convolutional neural networks Reliability

来源：评论

学校读者我要写书评

暂无评论

Descriptive and Predictive Analytics of Energy Security in India 2

Descriptive and Predictive Analytics of Energy Security in I...

引用

2nd IEEE International Conference on Advances in Computing, Communication and Applied informatics, ACCAI 2023

作者： Justus, J. Jean Mannish, S. St. Joseph's College of Engineering Computer Science and Engineering Tamil Nadu Chennai India St. Joseph's College of Engineering Dept. of Cse Tamil Nadu Chennai India

ISBN: (纸本)9798350315905

Energy security for a country is essential as energy management is crucial for a country's endurance and progress. For every development scheme, growth in the economy, or improving the exports to other countries in trade the most fundamental requirement is to match the energy production of the country to that of the growth of the country. This action is extremely crucial as it forms a basis where growth efficiently takes place. Another important aspect of energy security in today's age is overcoming the dependence on non-renewable forms of energy such as coal or natural gas. There are two prime reasons why this is significant. One is that these types of energy will not be exhausted in the upcoming years. Another reason is the long-term consequence on the environment producing and utilizing this energy conceives. Since this requires infrastructure and planning it is necessary to begin scheming on the renewable aspect. Therefore, it is imperative to focus on energy management on one hand and on the other focus on moving towards energy generation from non-renewable sources of energy to renewable sources. Furthermore, energy distribution management will benefit from the inequitable distribution of energy within a country. © 2023 IEEE.

关键词： Energy security

来源：评论

学校读者我要写书评

暂无评论

Prediction and Modelling of Suicidal Mental Tendencies

Prediction and Modelling of Suicidal Mental Tendencies

引用

IEEE Distributed Computing, VLSI, Electrical Circuits and Robotics (DISCOVER)

作者： Sunitha Guruprasad Arnav G Rao Shreeraksha Shivani K Varsha S N Yashoda K Dept. of Computer Science & Engineering St Joseph Engineering College Mangaluru India Visvesvaraya Technological University Belagavi Karnataka India Dept. of Computer Science & Engineering Ramaiah Institute of Technology Bengaluru India

ISBN: (数字)9798350350593

ISBN: (纸本)9798350350609

Suicide remains a pressing public health concern worldwide, necessitating effective predictive and preventive strategies. Predicting suicidal tendencies is a complex endeavor, as it involves multifaceted interactions between biological, psychological, and environmental factors. Advances in computational modeling and machine learning have provided promising avenues for understanding and forecasting these tendencies. Such models often integrate clinical data, including psychiatric history, symptom severity, and demographic information, with biological markers and psychosocial variables. The objective of our work is to identify patterns and risk factors that may indicate heightened suicidal risk. Ongoing research seeks to refine predictive models, incorporate real-time monitoring, and develop targeted interventions, aiming to reduce the incidence of suicide and provide timely support to those at risk. The application of suicide text analysis extends to user-friendly interfaces, allowing individuals to input text, receive predictions, and access relevant support. The analysis involves preprocessing techniques such as text normalization, stop-word removal, and stemming to enhance the quality of input data. Various machine learning algorithms, ranging from Naive Bayes to sophisticated ensemble methods, are explored to determine their effectiveness in predicting suicidal ideation.

关键词： Text analysis Computational modeling Soft sensors Predictive models Very large scale integration Natural language processing Real-time systems Bayes methods Wearable devices Robots

来源：评论

学校读者我要写书评

暂无评论

Adversarial Machine Learning Attack on Machine Learning-based Controller for Solar Inverters

Adversarial Machine Learning Attack on Machine Learning-base...

引用

IEEE Energy Conversion Congress and Exposition, ECCE

作者： Joaquin Massa Kyoung-Tak Kim Taesic Kim BoHyun Ahn Kumar Venayagamoorthy Jianwu Zeng Dept. of Electrical Engineering and Computer Science Texas A&M University-Kingsville Kingsville USA Dept. of Electrical Engineering Soongsil University Seoul Republic of Korea Dept. of Electrical Engineering and Computer Science University of Missour-Columbia Columbia USA Dept. of Electrical and Computer Engineering Clemson University Clemson USA Dept. of Electrical and Computer Engineering and Technology Minnesota State University Mankato Mankato USA

ISBN: (数字)9798350376067

ISBN: (纸本)9798350376074

Adversarial machine learning (ML) attacks are stealthy attacks designed to mislead the ML model results. This paper explores adversarial ML attacks that generate adversarial noisy input data in an ML-based controller in a solar inverter. Three types of ML models, long short-term memory (LSTM), gated recurrent unit (GRU)), and bidirectional-LSTM (Bi-LSTM), are designed to replace proportional-integral (PI) controller-based vector control for a solar inverter and two white-box adversarial ML attacks (Basic Iterative Method (BIM) attack and Fast Sign-Gradient Method (FGSM)) are applied to the ML controllers. It is observed that the adversary ML attacks designed in stealthy way do not affect the PI controller, while significantly degrading performance of the ML-based controllers. Moreover, the BIM attack is more effective than FGSM and Bi-LSTM-based controller is relatively robust to the attacks compared to peer.

关键词： PI control Inverters Adversarial machine learning Hardware Robustness Real-time systems Noise measurement Long short term memory Machine vector control Microprogramming

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：