检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Patra, Aswini Kumar Sahoo, Lingaraj Dept. of Computer Science and Engineering North Eastern Regional Institute of Science and Technology Itanagar India Dept. of Bio-Science and Bio-Engineering Indian Institute of Technology Guwahati India

Early identification of drought stress in crops is vital for implementing effective mitigation measures and reducing yield loss. Non-invasive imaging techniques hold immense potential by capturing subtle physiological changes in plants under water deficit. Sensor based imaging data serves as a rich source of information for machine learning and deep learning algorithms, facilitating further analysis aimed at identifying drought stress. While these approaches yield favorable results, real-time field applications requires algorithms specifically designed for the complexities of natural agricultural conditions. Our work proposes a novel deep learning framework for classifying drought stress in potato crops captured by UAVs in natural settings. The novelty lies in the synergistic combination of a pre-trained network with carefully designed custom layers. This architecture leverages the pre-trained network’s feature extraction capabilities while the custom layers enable targeted dimensionality reduction and enhanced regularization, ultimately leading to improved performance. A key innovation of our work involves the integration of Gradient-Class Activation Mapping (Grad-CAM), an explainability technique. Grad-CAM sheds light on the internal workings of the deep learning model, typically referred to as a"black box." By visualizing the model’s focus areas within the images, Grad-CAM fosters interpretability and builds trust in the model’s decision-making process. Our proposed framework achieves superior performance, particularly with the DenseNet121 pre-trained network, reaching a precision of 97% to identify the stressed class with an overall accuracy of 91%. Comparative analysis of existing state-of-the-art object detection algorithms reveals the superiority of our approach in significantly higher precision and accuracy. Thus, our explainable deep learning framework offers powerful approach for drought stress identification with high accuracy and actionable insights. Cop

关键词： Cams

来源：评论

学校读者我要写书评

暂无评论

Comprehensive Analysis of Machine Learning Approaches for Phishing Website Detection

Comprehensive Analysis of Machine Learning Approaches for Ph...

引用

Cognitive Computing in engineering, Communications, sciences and Biomedical Health Informatics (IC3ECSBHI), International Conference on

作者： Pragna K P Suhani Shetty Ashwini Kodipalli Ushasree. A Shridhar B. Devamane Roshan Joy Martis Dept. of Electronics and Communication Engineering Global Academy Of Technology Bengaluru Karnataka India Dept. of Artificial Intelligence & Data Science Global Academy of Technology Bengaluru Karnataka India Dept. of Computer Science & Engineering Gokaraju Lailavathi Engineering College Hyderabad India Department of Computer Science School of Computing Amrita Vishwavidyapeetam Mysuru Dept. of Electronics & Communication Engineering Manipal Institute of Technology Bengaluru Manipal Academy of Higher Education Manipal India

ISBN: (数字)9798331518523

ISBN: (纸本)9798331518530

Website phishing, a notable cybercrime, involves deceiving individuals by impersonating legitimate websites. URL spoofing, a common technique in website phishing, involves creating website names. The primary objective of this paper is to accurately differentiate between legitimate and phishing sites using relevant URL attributes and different binary classification machine learning algorithms. This study compares simple, bagging, and boosting algorithms across performance analysis metrics. The best-performing models are selected for an ensemble approach. Additionally, various hyperparameter tuning techniques are assessed for the Random Forest Classifier, and Explainable AI methods like LIME and SHAP are used to interpret findings. The experiment yields optimal parameters, achieving 89. % accuracy with an ensemble model with K-Nearest Neighbors, Decision Tree, XG Boost, and Cat Boost. These findings highlight the significance of ensemble models and hyperparameter tuning in machine learning.

关键词： Uniform resource locators Machine learning algorithms Explainable AI Phishing Computational modeling Biological system modeling Nearest neighbor methods Performance analysis Tuning Random forests

来源：评论

学校读者我要写书评

暂无评论

Explainable AI for Enhanced Intrusion Detection Systems

Explainable AI for Enhanced Intrusion Detection Systems

引用

Advancement in Technology (ICONAT), International Conference for

作者： G Ananya N Siri Ashwini Kodipalli Trupthi Rao Dept. of Computer Science and Engineering Global Academy of Technology Bengaluru Karnataka India Dept. of Electronic and Communication Engineering Global Academy of Technology Bengaluru Karnataka India Dept. of Artificial Intelligence & Data Science Global Academy of Technology Bengaluru Karnataka India

ISBN: (数字)9798350354171

ISBN: (纸本)9798350354188

As everything is turning digital there is a hike in the invasion of digital space, because of which advanced tools are needed. One such tool is the Intrusion Detection System (IDS). In this work, we have tried to detect and classify suspicious connections accurately. This is done by incorporating advanced computing into Intrusion Detection Systems. To achieve this, we have used the Coburg Intrusion Detection dataset. To achieve this, we used the Coburg intrusion detection dataset. We investigate the effectiveness of Single Classifiers and Bagging and Boosting in this context. We proposed a four-stage Ensemble model to classify the connections using its attributes. The proposed model uses four Single Classification models: Logistic Regression, KNN, Decision Tree, and SVM. Before applying the algorithms, the data was Preprocessed. Later, the models were hyper-tuned to increase the accuracy. The results showed that Random Forest outperforms other single classifiers and machine learning algorithms, with a mean accuracy of $\mathbf{9 9. 9 9 \%}$, a precision of $\mathbf{9 9. 9 9 \%}$, and an F1-score of $\mathbf{9 9. 9 9 \%}$. Furthermore, as models are usually hard to understand by using Explainable AI we can increase the transparency of the model. Explainable AI techniques are used to make the advanced computing models reliable and more understandable. This study aims to develop advanced and dependable intrusion detection systems

关键词： Support vector machines Machine learning algorithms Accuracy Explainable AI Computational modeling Intrusion detection Nearest neighbor methods Turning Reliability Random forests

来源：评论

学校读者我要写书评

暂无评论

Early Thermal Forest Fire Detection using UAV and Saliency map 5

Early Thermal Forest Fire Detection using UAV and Saliency m...

引用

5th International Conference on Contemporary Computing and Informatics, IC3I 2022

作者： Dilli, B. Suguna, M. Thiagarajar College of Engineering Dept. of Mechanical Engineering Tamil Nadu Madurai India Thiagarajar College of Engineering Dept. of Computer Science Engineering Tamil Nadu Madurai India

ISBN: (纸本)9798350398267

Wildfire is a dangerous disaster that threatens not only animals but also humans. Traditionally, Ground crew inspections form the basis for firefighting monitoring systems, which have several limitations. With the help of UAVs, before a wildfire becomes out of control, it is simple to spot it in its early stages. The last few years have seen a surge in interest in autonomous wildfire early detection using various deep learning algorithms based on UAV-based visual data. But the area of night vision surveillance is unattended. In this research, we concentrate on detecting the early stage of wildfires using YOLO model, a DL based object detection library on thermal images from UAVs and also the drawbacks in using the thermal images are solved using saliency maps fused with thermal images, which is then compared with the benchmark of the YOLO model trained only on thermal images. The proposed methodology is found to be capable enough to provide technical support in night surveillance to reduce the catastrophic losses of wild forest fires in their early stages and in terms of forest resources, and human and animal lives. © 2022 IEEE.

关键词： Unmanned aerial vehicles (UAV)

来源：评论

学校读者我要写书评

暂无评论

BSCIAM: A Blockchain based Secure Cloud Identity and Access Management Framework

BSCIAM: A Blockchain based Secure Cloud Identity and Access ...

引用

IEEE Students' Conference on Electrical, Electronics and computer science (SCEECS)

作者： Swatisipra Das Mohammad Sahil Naveen Kumar Pandit Rojalina Priyadarshini Sarada Prasad Gochhayat Postgraduate Dept. of Computer Science Fakir Mohan University Odisha India Dept. of Computer Science and Engineering C.V. Raman Global University Odisha India Dept. of Computer Science and Engineering Indian Institute of Technology Jammu J&K India

Cloud computing enables on-demand computation on remote servers and computers. Thanks to the adaptability and scalability of the infrastructure for data storage, processing, and management. To solve the security problems arising in cloud identity management techniques such as, dependency on a third-party token provider, leakage of user’s details, single point of failure, etc., decentralized cloud identity management systems came into the picture. Blockchain is a decentralized database providing immutability and transparency to recorded transaction data. The current token-based decentralized cloud identity management systems have limitations, including the absence of data access management procedures and a lack of security features. This paper suggests a Blockchain based Secure Cloud Identity and Access Management (BSCIAM) model for secure identity and access management, utilizing encryption and key sharing techniques. The proposed model has been implemented using Ethereum smart contracts for token-based identity management.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Comprehensive Analysis of Machine Learning Techniques in Autism Spectrum Disorder Detection

Comprehensive Analysis of Machine Learning Techniques in Aut...

引用

Distributed Computing and Electrical Circuits and Electronics (ICDCECE), IEEE International Conference on

作者： Lakshaya Kaushik Ranjit Singh Manjit Singh Dept. of Computer Science & Engineering Chandigarh University Gharuan Mohali Punjab India

ISBN: (数字)9798331533663

ISBN: (纸本)9798331533670

This review investigates the application of machine learning for (ASD) Autism Spectrum Disorder detection by reviewing existing models, discussing challenges, identifying future pathways for increasing detection and intervention services. We conducted a systematic review of studies from 2015 to 2024 that use ML techniques, including analysis of deep learning, traditional algorithms, and hybrid models, for the ASD detection, involving analyses of neuroimaging, behavioral, genetic, and physiological data. CNNs excel in neuroimaging, while RNNs aid time series behavioural data. Overall, the mutlimodal approach was more accurate that models reliant on single sources, ranging from 85% to 98% accurate. The two main issues were heterogeneity of data and limited feature selection. There is evidence in the literature, namely, that ML techniques may enhance ASD monitoring and identification. The future work should focus on creating datasets that consistently contain people from various backgrounds, creating interpretable machine learning models, and conducting clinical on these techniques.

关键词： Neuroimaging Autism Accuracy Systematics Time series analysis Medical services Genetics Biological neural networks Standards Systematic literature review

来源：评论

学校读者我要写书评

暂无评论

Applications of AI in Preservation of Kumaoni Culture Through Digital Archival 3

Applications of AI in Preservation of Kumaoni Culture Throug...

引用

3rd International Conference on Disruptive Technologies, ICDT 2025

作者： Dev, Vaishali Dhingra, Ansh Sharma, Shashi Kumar Koturiya, Nishant Graphic Era Hill University Dept. of Computer Science and Engineering Bhimtal India

ISBN: (纸本)9798331519582

Culture represents a society with human values and ethics. Without civilization there can be no culture. Thus it is very important to preserve our culture as it represents our values and ethics. Each region has its own characteristics culture which separates it from other regions not just geographically but in terms of traditions, art and literature. The Kumaon region in the Indian state of Uttarakhand lies in the eastern part of the state. It is blessed with a rich culture heritage and nature. The preservation of culture is done through traditional methods. The focus of this study is to use modern Artificial Intelligence techniques to achieve the preservation of culture. The Artificial Intelligence techniques which are used in this paper range from Convolutional Neural Networks, augmentation of data and fine tuning of models. To preserve it is necessary to archive the cultural elements so that it can be easily accessible in the future. The archival process can help in cataloguing the cultural items. By culture we mean traditional arts, such as Aipan, local literature texts and historical structures. The archival process seamlessly integrates the Artificial Intelligence methods to effectively preserve the cultural elements. This process can be further enhanced by fine tuning the Convolutional Neural Networks. This makes the model more robust and accurate than the existing model. The dataset quality can be further improved by using data augmentation techniques. This allows to generate more data from existing dataset. However, with the process comes some challenges like accountability and validity. This study explores these challenges which can help existing researchers in the field of culture and arts. © 2025 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

CAS-PUF: Current-Mode Array-Type Strong PUF for Secure Computing in Area Constrained SoCs

CAS-PUF: Current-Mode Array-Type Strong PUF for Secure Compu...

引用

Design, Automation and Test in Europe Conference and Exhibition

作者： Dimosthenis Georgoulas Yiorgos Tsiatouhas Vasileios Tenentes Dept. of Computer Science and Engineering VCAS Lab University of Ioannina Ioannina Greece

ISBN: (数字)9783982674100

ISBN: (纸本)9798331534646

Secure computing necessitates the integration in Systems-on-Chips (SoCs) of strong Physical Unclonable Functions (PUFs) that can generate a vast amount of Challenge Response Pairs (CRPs) for cryptographic keys generation, identification and authentication. However, the excessive area cost of strong PUF designs imposes integration difficulties to SoCs of area constrained applications, such as the IoT and mobile computing. In this paper, we present a novel strong PUF design, with silicon area requirements significantly lower than those of previous strong PUFs. The proposed Current-mode Array-type Strong PUF (CAS-PUF) is based on a current source topology of only six minimum size transistors, which is tolerant to power supply variation for enhanced reliability. Compared to previous strong PUFs, the CAS-PUF achieves the same number of CRPs with 20% to 72% less area size; while for the same area size, it provides 19 to 53 orders of magnitude higher number of CRPs. Furthermore, extensive Monte Carlo simulations on CAS-PDF show a reliability of 96.45% under ±10% power supply fluctuation; and 97.69% under temperature variation (0°C to 80°C), with an average uniqueness and uniformity of 50.01% and 49.54%, respectively. Therefore, CAS-PUF can be used as a hardware root of trust mechanism to secure computing in area constrained SoCs.

关键词： Temperature measurement Temperature sensors Power supplies Simulation Physical unclonable function Reliability engineering Silicon Topology Transistors Integrated circuit reliability

来源：评论

学校读者我要写书评

暂无评论

Leveraging Artificial Intelligence for Colon Cancer: Trends, Challenges and Innovations

Leveraging Artificial Intelligence for Colon Cancer: Trends,...

引用

Automation and Computation (AUTOCOM), International Conference on

作者： Sakshi Takkar Manik Rakhra Dept. of Computer Science and Engineering Lovely Professional University Phagwara Punjab India

ISBN: (数字)9798331542375

ISBN: (纸本)9798331542382

Artificial Intelligence (AI) has made huge enhancements in colon cancer research through its methods for diagnosis, treatment, and custom detection of one of the most common and fatal cancers suffered globally. Through AI deep learning and machine learning methods, tremendous possibilities have been realized in automating, for example, polyp detection in colonoscopy images, cancer classification from histopathological slides, and even predicting disease treatment outcomes. The current direction is towards transfer learning, hybrid model building and AI applications in surgery robots aimed at more accurate diagnosis and treatment procedures. Besides them, however, they have serious disadvantages, access to computer learning systems and datasets are limited, AI models come without explanation, embedding of artificial intelligence into clinical practices becomes troublesome. In future development, emphasize should be on creating databases, explainable ethics AI approaches, building bridges between developers and practitioners in medicine. The objective of the paper is how to incorporate research that has been done in moving colon cancer AI systems into practice, understanding the AI satellites put on colon cancer research.

关键词： Deep learning Ethics Technological innovation Computational modeling Buildings Transfer learning Surgery Colorectal cancer data models Artificial intelligence

来源：评论

学校读者我要写书评

暂无评论

Arrhythmia Classification Using Graph Neural Networks Based on Correlation Matrix

Arrhythmia Classification Using Graph Neural Networks Based ...

引用

IEEE International Conference on Bioinformatics and Biomedicine (BIBM)

作者： Seungwoo Han Dept. of Electrical Engineering and Computer Science Tokyo University of Agriculture and Technology

ISBN: (数字)9798350386226

ISBN: (纸本)9798350386233

With the advancements in graph neural network (GNN), there has been increasing interest in applying GNN to electrocardiogram (ECG) analysis. In this study, we generated an adjacency matrix using correlation matrix of extracted ECG features and applied a GNN to classify arrhythmias. Our work was compared with existing approaches from the literature. The results demonstrated that precision and recall for all arrhythmia classes exceeded 50%, suggesting that proposed method can be considered an approach for arrhythmia classification.

关键词： Correlation coefficient Correlation Arrhythmia Biological system modeling Frequency-domain analysis Electrocardiography Feature extraction Graph neural networks Bioinformatics

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：