检索结果-内蒙古大学图书馆

41st International Conference on Machine Learning, ICML 2024

作者： Zhang, Jie Ma, Xiaosong Guo, Song Li, Peng Xu, Wenchao Tang, Xueyang Hong, Zicong Department of Computing The Hong Kong Polytechnic University Hong Kong Department of Computer Science and Engineering The Hong Kong University of Science and Technology Hong Kong School of Cyber Science and Engineering Xi'an Jiaotong University Shaanxi China

Fine-tuning the learnable prompt for a pre-trained vision-language model (VLM), such as CLIP, has demonstrated exceptional efficiency in adapting to a broad range of downstream *** prompt tuning methods for VLMs do not distinguish spurious features introduced by biased training data from invariant features, and employ a uniform alignment process when adapting to unseen target *** can impair the cross-modal feature alignment when the test data significantly deviate from the distribution of the training data, resulting in a poor out-of-distribution (OOD) generalization *** this paper, we reveal that the prompt tuning failure in such OOD scenarios can be attribute to the undesired alignment between the textual and the spurious *** a solution, we propose CoOPood, a fine-grained prompt tuning method that can discern the causal features and deliberately align the text modality with the invariant ***, we design two independent contrastive phases using two lightweight projection layers during the alignment, each with different objectives: 1) pulling the text embedding closer to the invariant image embedding and 2) pushing the text embedding away from the spurious image *** have illustrated that CoOPood can serve as a general framework for VLMs and can be seamlessly integrated with existing prompt tuning *** experiments on various OOD datasets demonstrate the performance superiority over state-of-the-art methods. Copyright 2024 by the author(s)

关键词： Visual languages

来源：评论

学校读者我要写书评

暂无评论

Big Data Analysis of Cognitive Cloud Computing Based Intelligent Healthcare System 10

Big Data Analysis of Cognitive Cloud Computing Based Intelli...

引用

10th IEEE Uttar Pradesh Section International Conference on Electrical, Electronics and computer engineering, UPCON 2023

作者： Shah, Nirav Saxena, Aakansha Kumar, Yogesh Silveroak University Department of Cse Ahmedabad India School of It Artificial Intelligence & Cyber Security Rashtirya Raksha University India Pandit Deendayal Energy University Department of Computer Science and Engineering Gujarat Gandhinagar India

ISBN: (纸本)9798350382471

The medical sector domain is one of the most rapidly rising and demanding fields in the current era. With the increased usage of technology and the internet, it has become critical for healthcare professionals to adapt and exploit digital technologies in order to stay up with changing patient needs. Big Data can revolutionize the domain of healthcare by providing significant improvements in disease diagnosis, treatment planning, and drug development. Cognitive cloud computing-based intelligent healthcare systems can provide a comprehensive solution to the medical sector processing and analyzing large amounts of healthcare data. This study investigates the role, advantages, and disadvantages of using Big Data in the medical industry and outlines the structure of an intelligent healthcare system using cognitive cloud computing. In addition to this, the contribution of the researchers in this particular realm has been also discussed in detail. © 2023 IEEE.

关键词： Artificial Intelligence Big Data Cognitive Cloud Computing Deep Learning Healthcare Machine Learning

来源：评论

学校读者我要写书评

暂无评论

DCVCP: Experimental Evaluation of Leukaemia Disease Prediction Using Deep Convoluted Vector Classification Principle

DCVCP: Experimental Evaluation of Leukaemia Disease Predicti...

引用

2024 International Conference on Innovative Computing, Intelligent Communication and Smart Electrical Systems, ICSES 2024

作者： Balachander, Bhuvaneswari Sivamurugan, S. Gomathi, M. Jayanthi, J. Adkane, R.V. Kanan, M. Dr.M.G.R Educational and Research Institute Department of Electronics and Communication Engineering Chennai India Sri Sairam Engineering College Department of Artificial Intelligence and Data Science Chennai India Dhanalakshmi College of Engineering Department of Cyber Security Chennai India Sona College of Technology Department of Computer Science and Engineering Salem India Yeshwantrao Chavan College of Engineering Department of Mechanical Engineering Nagpur India College of Engineering University of Business and Technology Department of Industrial Engineering Jeddah21448 Saudi Arabia College of Engineering Zarqa University Department of Mechanical Engineering Zarqa Jordan

ISBN: (纸本)9798331543617

A wide variety of cancers, including skin, breast, lung, and blood malignancies such as lymphoma and leukemia, are defined by the uncontrolled growth of cells within the body. Acquired lymphoblastic leukemia is among the most prominent cancers. Blood cancer diagnoses are notoriously time-consuming because hematologists often make mistakes. With the use of cutting-edge tools like Deep Learning and Machine Learning, this study considers a fresh approach to leukemia classification. The precise diagnosis of blood cancer is a crucial medical process, but it is also one of the most costly and challenging for clinical staff to do. Blood cancer is one of many severe forms of the disease. The devastating and deadly blood cancers, lymphoma and leukemia, impact people of all ages and genders, and they are a leading cause of other illnesses and an elevated survival rate. Damage to and an increase in immature monocytes, neutrophils, eosinophils, and lymphocytes are hallmarks of both leukemia and lymphoma. Predicting and treating blood cancer early is crucial for improving survival rates in the health industry. These days, a number of manual methods exist for analyzing and predicting blood cancer based on microscopic medical reports of white blood cell images;this method is both reliable and deadly. It takes a lot of time and effort to manually forecast and analyze eosinophils, lymphocytes, monocytes, and neutrophils. The Deep Convoluted Vector Classification Principle (DCVCP) is a new deep learning technique for leukemia illness detection that was described in this study. To test how well it works, it was cross-validated with the current method, the Convolutional Neural Network (CNN). findings from this study show that DCVCP can effectively leverage data from network sensors to identify blood cancer, which might lead to faster findings and lower costs in clinical settings. © 2024 IEEE.

关键词： Lung cancer

来源：评论

学校读者我要写书评

暂无评论

Detecting Hydrocephalus Using CT Scans and Machine Leaning Techniques 2

Detecting Hydrocephalus Using CT Scans and Machine Leaning T...

引用

2nd International engineering Conference on Electrical, Energy, and Artificial Intelligence, EICEEAI 2023

作者： Pourhossain Latifi, Muhammad Sadegh Gourabi, Ahmad Alqammaz, Abdullah Moradi, Ehsan Hosseini, Rahil Boostani, Reza Sabeti, Maliheh Islamic Azad University North Tehran Branch Department of Computer Engineering Tehran Iran Zarqa University Faculty of Information Technology Department of Cyber Security Zarqa Jordan College of Engineering University of Business and Technology Jeddah21448 Saudi Arabia Shahid Beheshti University of Medical Sciences Department of Neurosurgery Tehran Iran Islamic Azad University Department of Computer Engineering Shahr-e-Qods Branch Tehran Iran Shiraz University Faculty of Electrical and Computer Engineering Cse and It Department Shiraz Iran

ISBN: (纸本)9798350373363

Every year, hydrocephalus affects many people, whether they are infants or adults with higher incidence in pediatric group. This means that a large number of families affected by this disease, and automatic determination of hydrocephalus in medical images could help people who are affected with this disease and their condition is at a dangerous stage. In this research, we try to implement an algorithm that could detect hydrocephalus from CT images. Our work is divided into two stages of implementing the proposed algorithm and comparing it with the results of similar researches. We try to design and implement all steps from scratch. The proposed model reaches the result with appropriate accuracy but slower running time. Our Method is able to detect hydrocephaly from 40 patients CT Images and calculate ventricular indices with accuracy of 64% to 93% in comparison with manually extracted ventricular indices by a neurosurgeon. In addition, we also checked the K-nearest neighbor (KNN) and Watershed methods and concluded that Region Growing is the best among this category. © 2023 IEEE.

关键词： computerized tomography

来源：评论

学校读者我要写书评

暂无评论

Facial Recognition Utilizing the CLM Model and Supervised Machine Learning: A Comparative Analysis 25

Facial Recognition Utilizing the CLM Model and Supervised Ma...

引用

25th International Arab Conference on Information technology, ACIT 2024

作者： Iqtait, Musab Ababneh, Jafar Abu-Jassar, Amer Al-Refai, Mohammad Dmour, Mohammad Arabiat, Mohammad Binsaddig, Ruaa Zarqa University Faculty of Information Technology Department of Artificial Intelligence Zarqa Jordan Zarqa University Faculty of Information Technology Cyber Security Department Zarqa Jordan Amman Arab University University College of of Computer Sciences and Information Amman Jordan Zarqa University Faculty of Information Technology Software Engineering Department Zarqa Jordan University of Business and Technology College of Business Administration Department of Finance Jeddah21432 Saudi Arabia

ISBN: (纸本)9798331540012

Facial recognition is a common challenge in artificial intelligence. We extensively utilized this program in our daily life. Face recognition software was implemented on many phones to protect user identities and promptly recognize Facebook users in photographs. Despite the numerous methodologies proposed to date, the actual implementation of facial recognition continues to pose significant challenges. Distinguishing individuals by a fundamental method relies on various factors, including lighting, postural differences, and partial facial obstruction. This study aims to develop a facial recognition technique utilizing a constrained local model (CLM) and a machine learning algorithm. It is evaluated with KNN, SVR, and CCA. Furthermore, it has employed CLM and CCA to achieve a maximum recognition accuracy of 3.98 MAE. © 2024 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Multi-Behavioral Recommender System with Heterogeneous Graph Learning

Multi-Behavioral Recommender System with Heterogeneous Graph...

引用

Consumer Electronics and computer engineering (ICCECE), 2021 IEEE International Conference on

作者： Jiaming Feng Yang Liu Junjie Zhang School of Computer Science and Engineering (School of Cyber Security) University of Electronic Science and Technology of China Chengdu China

With the rapid development of online platforms and the explosive growth of user-generated data, personalized recommendation systems have become increasingly crucial in delivering relevant content to users. Traditional recommendation approaches often focus on modeling user-item interactions, neglecting the rich contextual information associated with user behaviors. In this paper, we propose a novel multi-behavior recommendation system based on heterogeneous graph learning. By leveraging the power of heterogeneity, our model integrates various types of user behaviors, such as browsing, adding cart, and purchasing, into a unified framework. We design a graph-based representation for capturing the complex relationships between users, items, and behaviors. Additionally, we introduce a graph neural network approach to effectively propagate and aggregate information across the heterogeneous graph. Experimental results on a real-world dataset demonstrate that our proposed system outperforms state-of-the-art recommendation methods in terms of accuracy and diversity. The findings highlight the potential of heterogeneous graph learning in enhancing the performance of multi-behavior recommendation systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Malware Detection and Classification for Internet of Things Using Self-Attention Based Long Short-Term Memory 3

Malware Detection and Classification for Internet of Things ...

引用

3rd IEEE International Conference on Distributed Computing and Electrical Circuits and Electronics, ICDCECE 2024

作者： Sudhakar, K. Husain, Saif O. Babu, D. Raja Kumar, M. Ashok Dineshkumar, R. Nitte Meenakshi Institute Of Technology Dept. Of Ai & Ds Bengaluru India College Of Technical Engineering The Islamic University Department Of Computers Techniques Engineering Najaf Iraq Sr University Department Of Electrical And Electronics Engineering Warangal India Skyline University Nigeria Department Of Computer Science Software Engineering And Cyber Security Kano State Nigeria Saveetha School Of Engineering Saveetha Institute Of Medical And Technical Sciences Saveetha University Department Of Electronics And Communication Engineering Chennai India

ISBN: (纸本)9798350318609

The Internet of Things (IoT) connects numerous intelligent devices providing security features that interact with default settings accessed through applications. Additionally, Deep Learning (DL)-based mechanisms was involved for malware detection to identify and mitigate attacks. The Adaptive Synthetic Sampling (ADASYN) technique addresses class imbalances, especially when splitter class instances are distributed data in complex patterns. The proposed Self-Attention based Long Short-Term Memory (LSTM) classification method facilitates the handling of long-range dependencies in transferred data sequences offering high flexibility and adaptability to varying input lengths. For feature selection Particle Swarm Optimization (PSO) is utilized iteratively optimizes a solution in a high-dimensional space although it may have a low convergence rate due to its iterative process. The obtained results demonstrate the proposed Self-Attention based LSTM model achieves better accuracy reaching 98.05% on the NSL-KDD dataset and 99.01 % on the UNSW-NB15 dataset. These results ensure accurate classification compared to other existing methods, such as Convolutional Neural Network (CNN) and Recurrent Neural Network (RNN). © 2024 IEEE.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Dynamic Graph-Enhanced Sequential LSTM for Context-Aware Point-of-Interest Recommendation

Dynamic Graph-Enhanced Sequential LSTM for Context-Aware Poi...

引用

International Conference on Wavelet Active Media technology and Information Processing (ICWAMTIP)

作者： Zhang Yuwen Ren Shaobo School of Computer Science and Engineering (School of Cyber Security) University of Electronic Science and Technology of China Chengdu China

ISBN: (数字)9798331519254

ISBN: (纸本)9798331519261

Next Point-of-Interest (POI) recommendation is a critical task in location-based services, aiming to predict the user's next POI based on historical check-in data. Existing approaches often rely on static graphs or sequential models to capture user preferences, which are limited by the sparsity of check-in data and the inability to dynamically model spatio-temporal transitions. To address these challenges, we propose the Dynamic Graph-Enhanced LSTM (DGEL) model for next POI recommendation. DGEL constructs dynamic POI transition graphs by combining Transformer-based graph learning with embedding techniques, capturing implicit relationships between POIs. Additionally, a Customized Sequential LSTM (CS-LSTM) is introduced to model long-term and short-term dependencies, incorporating spatial, temporal, and relational signals. Experimental results on Gowalla and Foursquare datasets demonstrate that DGEL achieves significant performance improvements over state-of-the-art baselines, highlighting its effectiveness in leveraging dynamic graphs and structured sequential learning for POI recommendation.

关键词： Computational modeling Neural networks Information processing Media Transformers Data models Long short term memory Context modeling

来源：评论

学校读者我要写书评

暂无评论

Hybrid of Distributed Cumulative Histograms and Classification Model for Attack Detection

引用

computer Systems Science & engineering 2023年第5期45卷 2235-2247页

作者： Mostafa Nassar Anas M.Ali Walid El-Shafai Adel Saleeb Fathi E.Abd El-Samie Naglaa F.Soliman Hussah Nasser AlEisa Hossam Eldin H.Ahmed Department of Electronics and Electrical Communications Engineering Faculty of Electronic EngineeringMenoufia UniversityMenouf32952Egypt Alexandria Higher Institute of Engineering&Technology(AIET) AlexandriaEgypt Department of Computer Science Security Engineering LaboratoryPrince Sultan UniversityRiyadh11586Saudi Arabia Department of Information Technology College of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia Department of Computer Sciences College of Computer and Information SciencesPrincess Nourah Bint Abdulrahman UniversityRiyadhSaudi Arabia

Traditional security systems are exposed to many various attacks,which represents a major challenge for the spread of the Internet in the *** techniques have been suggested for detecting attacks using machine learning and deep *** significant advantage of deep learning is that it is highly efficient,but it needs a large training time with a lot of ***,in this paper,we present a new feature reduction strategy based on Distributed Cumulative Histograms(DCH)to distinguish between dataset features to locate the most effective *** histograms assess the dataset instance patterns of the applied features to identify the most effective attributes that can significantly impact the classification *** different models for detecting attacks using Convolutional Neural Network(CNN)and Long Short-Term Memory Network(LSTM)are also *** accuracy test of attack detection using the hybrid model was 98.96%on the UNSW-NP15 *** proposed model is compared with wrapper-based and filter-based Feature Selection(FS)*** proposed model reduced classification time and increased detection accuracy.

关键词： Feature selection DCH LSTM CNN security systems

来源：评论

学校读者我要写书评

暂无评论

Inflammation Segmentation Module using Convolution Neural Network for Alzheimer's Disease Detection using Magnetic Resonance Imaging Inputs 9

Inflammation Segmentation Module using Convolution Neural Ne...

引用

9th International Conference on Science, technology, engineering and Mathematics, ICONSTEM 2024

作者： Shetty, Sahana Ramesh, S. Nithya, R. Thangaraj, P. Antony Vijay, J. Sangamithrai, K. School of Cse Fet - Jain Department of CSE-AIML And Cyber Security Bangalore India College of Engineering & Technology Srm Institute of Science And Technology Department of Computing Technologies Tamilnadu Kattankulathur India Bannari Amman Institute of Technology Department of Cse Tamilnadu Erode India Builders Engineering College Department of Cse Tamilnadu Tirupur India Karpagam College of Engineering Department of Cse Tamilnadu Coimbatore India Saveetha School of Engineering Saveetha Institute of Medical And Technical Sciences Saveetha University Department of Computer Science Chennai India

ISBN: (纸本)9798350365092

Magnetic Resonance Imaging (MRI) inputs have been most commonly used in diagnosing Alzheimer's Disease (AD) in recent years. The patterns of brain functions are deliberately observed in the MRI inputs for efficient AD detection. This article introduces an Inflammation Segmentation Module (ISM) using a convolution neural network (CNN) to augment better AD detection. This segmentation module identifies the textural differences in inflammation regions through training input cross-correlation. The cross-correlation process is reliable in identifying similar and differential textural features to improve the detection accuracy. The CNN is trained using an external dataset and correlated outputs for high precision. This proposed module is reliable in detecting inflammation textures with high precision and less analysis time. © 2024 IEEE.

关键词： Textures

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：