检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Chen, Jiahao Yan, Diqun Dong, Li Faculty of Electrical Engineering and Computer Science Ningbo University Zhejiang Ningbo315211 China College of Computer Science and Technology Zhejiang University Zhejiang Hangzhou310027 China Key Laboratory of Mobile Network Application Technology of Zhejiang Province Zhejiang Ningbo315211 China

The presence of adversarial examples poses a significant threat to deep learning models and their applications. Existing defense methods provide certain resilience against adversarial examples, but often suffer from decreased accuracy and generalization performance, making it challenging to achieve a trade-off between robustness and generalization. To address this, our paper interprets the adversarial example problem from the perspective of sample distribution and proposes a defense method based on distribution shift, leveraging the distribution transfer capability of a diffusion model for adversarial defense. The core idea is to exploit the discrepancy between normal and adversarial sample distributions to achieve adversarial defense using a pretrained diffusion model. Specifically, an adversarial sample undergoes a forward diffusion process, moving away from the source distribution, followed by a reverse process guided by the protected model (victim model) output to map it back to the normal distribution. Experimental evaluations on CIFAR10 and ImageNet30 datasets are conducted, comparing with adversarial training and input preprocessing methods. For infinite-norm attacks with 8/255 perturbation, accuracy rates of 78.1% and 83.5% are achieved, respectively. For 2-norm attacks with 128/255 perturbation, accuracy rates are 74.3% and 82.5%. Additional experiments considering perturbation amplitude, diffusion iterations, and adaptive attacks also validate the effectiveness of the proposed method. Results demonstrate that even when the attacker has knowledge of the defense, the proposed distribution-based method effectively withstands adversarial examples. It fills the gaps of traditional approaches, restoring high-quality original samples and showcasing superior performance in model robustness and generalization. Copyright © 2023, The Authors. All rights reserved.

关键词： Normal distribution

来源：评论

学校读者我要写书评

暂无评论

An Integrated Image Classification Approach to Detect the Flood Prone Areas using Sentinel-1 Images

An Integrated Image Classification Approach to Detect the Fl...

引用

International Conference on Computing for Sustainable Global Development (INDIACom)

作者： Mohammed Siddique Tasneem Ahmed Mohd Shahid Husain Department of Computer Application Advanced Computing Research Lab Integral University Lucknow India Faculty of Information Technology Majan University College Muscat Oman Department of Information Technology CAS-Ibri University of Technology and Applied Sciences Muscat Oman

The untimely monsoon floods and the rise in the level of rivers in the north Indian state of Uttar Pradesh (UP) is a major challenge for coordinating relief and rescue operations. To address this challenge, the experts that are deploying a variety of techniques to evaluate the damages might get useful information from synthetic aperture radar (SAR) images which offer images in close to real-time. The Sentinel-1A satellite images from the European Space Agency provide C-band SAR images that have helped researchers achieve enhanced processes of flood detection and flood mapping in flood-prone areas. In regular cases, the Random Forest (RF) and K-Nearest Neighbor (KNN) supervised classification techniques provide higher accuracies than the unsupervised K-Means classification techniques. The integrated approach with the resultant accuracy provides long-reliable elucidation towards flood damages and assessment control. Thenceforth, the RF and KNN supervised classifications based outcomes were integrated with the outcomes of K-Means iterative clustering unsupervised classification producing an enhanced overall accuracy. The overall accuracy from the integrated approach for the city of Basti and Ayodhya is 97% and 93.8% which is higher than the accuracies derived from the supervised and unsupervised classifications. Further, these processed results when refined by time-series analysis over a number of years along with implementing change detection techniques would help in monitoring the flood situation and undertaking precautionary measures.

关键词： Radio frequency Satellites Time series analysis Urban areas Vegetation mapping Radar polarimetry Rivers

来源：评论

学校读者我要写书评

暂无评论

Deep Learning Techniques for Advanced Large-Scale Data Processing

Deep Learning Techniques for Advanced Large-Scale Data Proce...

引用

International Conference on Computing and Networking technology (ICCNT)

作者： Mp Sunil Jyoti Shekhawat T. Vanaja Vazeer Ali Mohammed Shailesh J. Thaware Vaishali Singh Department of Electronics and Communication Engineering Faculty of Engineering and Technology JAIN (Deemed-to-be University) Bangalore Karnataka India Department of Computer Science & Application Vivekananda Global University Jaipur India Department of Information Technology Prince Shri Venkateshwara Padmavathy Engineering College Chennai Department of Computer & Information System Lewis University Illinois USA Department of Computer Engineering Vishwakarma Institute of Technology Pune INDIA Maharishi School of Engineering and Technology Maharishi University of Information Technology Uttar Pradesh India

ISBN: (数字)9798350370249

ISBN: (纸本)9798350370270

Deep getting to know techniques have become increasingly more popular due to their potential to improve huge-scale records processing performance. Leveraging specialized hardware which includes GPUs, deep getting to know algorithms can attain mind-blowing pace will increase compared to traditional algorithms. further, deep gaining knowledge of fashions can discover complicated non-linear relationships in big datasets, permitting greater accurate overall performance and less fake predictions. Latest advances in deep mastering have enabled it to be implemented to remedy a plethora of information processing duties, from photo recognition to natural language processing. CNNs, RNNs, and GANs are a number of the maximum famous deep gaining knowledge of fashions used in tackling complex data processing obligations. agencies consisting of Google, fb, and Amazon use these sophisticated frameworks to build predictive models that help improve numerous programs. Deep getting to know strategies offer exquisite ability to take massive-scale information processing to the following stage. corporations are increasingly more becoming privy to the emerging opportunities, and actively investing in deep mastering to liberate the untapped strength of their information.

关键词： Deep learning Accuracy Scalability Information processing Predictive models Data processing Prediction algorithms Hardware Natural language processing Optimization

来源：评论

学校读者我要写书评

暂无评论

Cryptocurrency and Data Privacy in Human Resource Management 3

Cryptocurrency and Data Privacy in Human Resource Management

引用

3rd International Conference on Advance Computing and Innovative Technologies in Engineering, ICACITE 2023

作者： Chandrasekeran, Indhumathi Dharmaraj, A. Juyal, Ashima Shravan, M. Deb Barman, Rajesh Lourens, Melanie Department of Commerce with Computer Application PSG College of Arts & Science CoimbatoreTamil Nadu India Department of Management Karpagam Academy of Higher Education CoimbatoreTamil Nadu India Uttaranchal Institute of Technology Uttaranchal University India School of Business SR University WarangalTelangana India Department of Commerce Bodoland University Kokrajhar BTR Assam783370 India Faculty of Management Sciences Durban University of Technology South Africa

ISBN: (纸本)9798350399264

Cryptocurrency and data privacy are two important considerations in human resource management. Cryptocurrency, such as Bitcoin, can be used to pay employees, provide incentives, and manage payroll, but it also poses risks such as price volatility and regulatory issues. Data privacy, on the other hand, is the protection of personal information from unauthorized access and use. The use of blockchain technology, which is the underlying technology behind many cryptocurrencies, can enhance data privacy in human resource management by providing a secure and tamper-proof record of all transactions. However, it's important to comply with all applicable laws and regulations related to the use of cryptocurrency and data privacy in human resource management. © 2023 IEEE.

关键词： Human resource management

来源：评论

学校读者我要写书评

暂无评论

Optimizing Drug Dosage and Medication Management for Geriatric Patients with Artificial Neural Networks

Optimizing Drug Dosage and Medication Management for Geriatr...

引用

International Conference on ICT in Business Industry & Government (ICTBIG)

作者： Beemkumar N Daljeet Pal Singh Zahid Ahmed Department of Mechanical Engineering Faculty of Engineering and Technology JAIN (Deemed-to-be University) Bagalore Maharishi School of Engineering & Technology Maharishi University of Information Technology Uttar Pradesh India Department of Computer Science & Application Vivekananda Global University Jaipur India

In this paper, we utilize ANNs to investigate how medicine dosing and management might be improved for the elderly population. The suggested strategy calls for the use of patient-specific data as inputs for the ANNs, such as age, renal and hepatic function, body mass index, and preexisting prescription profiles. The neural network model dynamically changes drug doses using the combination of the Multilayer Perceptron (MLP) algorithm, Backpropagation, and the Stochastic Gradient Descent (SGD) algorithm to maximize therapeutic benefits and reduce the likelihood of adverse drug events. The research highlights the improved accuracy, sensitivity, specificity, precision, and F1 score of the suggested technique compared to numerous conventional methods. As a result, the suggested system has great potential to revolutionize medication management for the elderly population since it is more personalized, flexible, patient-safe, effective, and efficient in its use of resources.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Emerging Trends of Language Models and Their Significance in Digital Era

Emerging Trends of Language Models and Their Significance in...

引用

International Conference on Advances in Information technology (ICAIT)

作者： Rishi Kumar Praveen Gujjar J Guru Prasad M S Raghavendra M Devadas Lubna Ambreen Rahul Kumar Pal Department of Computer Science and Engg. Graphic Era (Deemed to be University) Dehradun India Faculty of Management Studies JAIN (Deemed-to-be University) Bengaluru India Department of Computer Science and Engineering Gitam School of Technology Gitam (Deemed-to-be University) Bengaluru India Department of Computer Application Graphic Era (Deemed to be University) Dehradun India

ISBN: (数字)9798350383867

ISBN: (纸本)9798350383874

This research investigates the rapidly evolving landscape of language models and their emerging trends. Language models, powered by innovative techniques in machine learning and natural language processing, have witnessed significant advancements in recent years. Language models have seen remarkable advancements in recent years, with emerging trends shaping the landscape of natural language processing (NLP). This research explores the latest developments in language models, such as transformer architectures and deep learning techniques, and assesses their significance in various applications. This research tries to understand the impact of these trends on tasks like text generation, sentiment analysis, and machine translation. This research also explores these advancements' potential societal and ethical implications, highlighting the need for responsible AI development. This study underscores the importance of staying updated on emerging trends in language models to harness their full potential while ensuring responsible and ethical usage in the ever-evolving field of NLP.

关键词： Training Ethics Data privacy Sentiment analysis Medical services Market research Transformers Natural language processing Data models Machine translation

来源：评论

学校读者我要写书评

暂无评论

Segment Together: A Versatile Paradigm for Semi-Supervised Medical Image Segmentation

引用

IEEE Transactions on Medical Imaging 2025年 PP卷 PP页

作者： Zeng, Qingjie Xie, Yutong Lu, Zilin Lu, Mengkang Wu, Yicheng Xia, Yong Northwestern Polytechnical University National Engineering Laboratory for Integrated Aero-Space-Ground-Ocean Big Data Application Technology School of Computer Science and Engineering Xi’an710072 China The University of Adelaide Australian Institute for Machine Learning AdelaideSA5000 Australia Monash University Faculty of Information Technology Department of Data Science and AI Australia

The scarcity of annotations has become a significant obstacle in training powerful deep-learning models for medical image segmentation, limiting their clinical application. To overcome this, semi-supervised learning that leverages abundant unlabeled data is highly desirable to enhance model training. However, most existing works still focus on specific medical tasks and underestimate the potential of learning across diverse tasks and datasets. In this paper, we propose a Versatile Semi-supervised framework (VerSemi) to present a new perspective that integrates various SSL tasks into a unified model with an extensive label space, exploiting more unlabeled data for semi-supervised medical image segmentation. Specifically, we introduce a dynamic task-prompted design to segment various targets from different datasets. Next, this unified model is used to identify the foreground regions from all labeled data, capturing cross-dataset semantics. Particularly, we create a synthetic task with a CutMix strategy to augment foreground targets within the expanded label space. To effectively utilize unlabeled data, we introduce a consistency constraint that aligns aggregated predictions from various tasks with those from the synthetic task, further guiding the model to accurately segment foreground regions during training. We evaluated our VerSemi framework against seven established SSL methods on four public benchmarking datasets. Our results suggest that VerSemi consistently outperforms all competing methods, beating the second-best method with a 2.69% average Dice gain on four datasets and setting a new state of the art for semi-supervised medical image segmentation. Code is available at https://***/maxwell0027/VerSemi. © 1982-2012 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Dimensionality Reduction Using UMAP and TSNE Technique

Dimensionality Reduction Using UMAP and TSNE Technique

引用

International Conference on Advances in Information technology (ICAIT)

作者： Mudit Mittal Praveen Gujjar J Guru Prasad M S Raghavendra M Devadas Lubna Ambreen Vikash Kumar Department of Computer Science and Engg. Graphic Era (Deemed to be University) Dehradun India Faculty of Management Studies JAIN (Deemed-to-be University) Bengaluru India Department of Computer Science and Engineering Gitam School of Technology Gitam (Deemed-to-be University) Bengaluru India Department of Computer Application Graphic Era (Deemed to be University) Dehradun India

ISBN: (数字)9798350383867

ISBN: (纸本)9798350383874

The increasing complexity and high dimensionality of datasets in various fields, such as genomics, image analysis, and natural language processing, have posed significant challenges for effective data visualization and analysis. Dimensionality reduction techniques play a crucial role in addressing these challenges by transforming high-dimensional data into lower-dimensional representations while preserving the inherent structure and relationships within the data. This research focuses on the application of two prominent dimensionality reduction techniques: Uniform Manifold Approximation and Projection (UMAP) and t-distributed Stochastic Neighbor Embedding (t-SNE). UMAP and t-SNE have gained widespread popularity for their ability to capture complex nonlinear relationships and preserve local and global structures in the data. This paper Deals with parameter-tuning strategies for both UMAP and t-SNE to optimize their performance based on different data characteristics. Practical guidelines are provided to aid researchers and practitioners in selecting appropriate dimensionality reduction techniques and parameter settings for specific applications. The findings of this research contribute to a deeper understanding of the capabilities and limitations of UMAP and t-SNE, providing valuable insights into their practical applicability in the analysis and visualization of high-dimensional datasets.

关键词： Dimensionality reduction Manifolds Iris Image analysis Data analysis Data visualization Genomics

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning for Predictive Modeling and Management of Rare Genetic Disorders in Pediatric Healthcare

Reinforcement Learning for Predictive Modeling and Managemen...

引用

International Conference on ICT in Business Industry & Government (ICTBIG)

作者： Pavan Chaudhary Inzimam R Hannah Jessie Rani Maharishi School of Engineering & Technology Maharishi University of Information Technology Uttar Pradesh India Department of Computer Science & Application Vivekananda Global University Jaipur India Electrical and Electronics Engineering Faculty of Engineering and Technology JAIN (Deemed-to-be University) Bangalore Karnataka

The difficulties caused by rare genetic abnormalities in pediatric healthcare highlight the need for advances in predictive modeling and treatment choices. An innovative method is discussed in the article “Reinforcement Learning for Predictive Modeling and Management of Rare Genetic Disorders in Pediatric Healthcare.” Deep Q-Networks (DQN), Proximal Policy Optimization (PPO), and Actor-Critical are the three RL algorithms used in the suggested approach. To determine the most effective treatment for children with uncommon genetic illnesses, we used these algorithms to create a prediction model. To improve healthcare quality, we carefully evaluate the suggested approach alongside six alternative methods with overlapping but distinct names. We also take accuracy and precision into account while making our assessments. The suggested strategy, which we refer to as the “Novel Genetic Predictor (NGP),” excels in every metric we tested. The results show a considerable improvement in the detection of rare genetic disorders in children. NGP also demonstrates an enhanced ability to make precise predictions, which opens the path for treatment alternatives that are both patient-specific and more effective. This research not only shows the promise of RL in pediatric healthcare, but also emphasizes the need for tailored care for children with rare genetic disorders. The results show that NGP is better than other methods, and they provide doctors a tool they may use to better care for their patients.

关键词：

来源：评论

学校读者我要写书评

暂无评论

3-D Near-Field Localization by Jointly Exploiting Spatial and Temporal Information Based on a Nonuniform Cross Array

3-D Near-Field Localization by Jointly Exploiting Spatial an...

引用

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Zelong Yi Hua Chen Zhiwei Jiang Wei Liu Qing Wang Gang Wang Faculty of Electrical Engineering and Computer Science Ningbo University Ningbo China Zhejiang Key Laboratory of Mobile Network Application Technology Ningbo China School of Electronic Engineering and Computer Science Queen Mary University of London London UK School of Electronic Informatin Engineering Tianjin University Tianjin China

In this paper, an underdetermined three-dimensional (3-D) near-field source localization method is proposed, based on a two-dimensional (2-D) symmetric nonuniform cross array. Firstly, the fourth-order cumulant of the near-field observations with multiple delay lags is exploited to construct virtual far-field pseudo-observations, leading to increased degrees of freedom (DOF); then, 2-D angles of the nearfield sources are jointly estimated by employing the recently proposed sparse and parametric approach (SPA) and Vandermonde decomposition technique, eliminating the need for parameter discretization. To estimate the range term, the one-dimensional (1-D) MUSIC algorithm is applied by resorting to the conjugate symmetry property of the signal's autocorrelation function. Numerical results are provided to demonstrate the superiority of our method.

关键词：

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：