检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Abdaljalil, Samir Kurban, Hasan Sharma, Parichit Serpedin, Erchin Atat, Rachad Electrical and Computer Engineering Texas A&M University College StationTX United States College of Science and Engineering Hamad Bin Khalifa University Doha Qatar Department of Computer Science Luddy School of Informatics BloomingtonIN United States Department of Computer Science and Mathematics Lebanese American University Beirut Lebanon

Large language models (LLMs) are increasingly deployed across diverse domains, yet they are prone to generating factually incorrect outputs—commonly known as "hallucinations." Among existing mitigation strategies, uncertainty-based methods are particularly attractive due to their ease of implementation, independence from external data, and compatibility with standard LLMs. In this work, we introduce a novel and scalable uncertainty-based semantic clustering framework for automated hallucination detection. Our approach leverages sentence embeddings and hierarchical clustering alongside a newly proposed inconsistency measure, SINdex, to yield more homogeneous clusters and more accurate detection of hallucination phenomena across various LLMs. Evaluations on prominent open- and closed-book QA datasets demonstrate that our method achieves AUROC improvements of up to 9.3% over state-of-the-art techniques. Extensive ablation studies further validate the effectiveness of each component in our framework. © 2025, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

MULTI-LF: A Unified Continuous Learning Framework for Real-Time DDoS Detection in Multi-Environment Networks

arXiv

引用

arXiv 2025年

作者： Rustam, Furqan Obaidat, Islam Jurcut, Anca Delia School of Computer Science University College Dublin Ireland Department of Computer Systems Technology North Carolina A&T State University NC United States

Detecting Distributed Denial of Service (DDoS) attacks in Multi-Environment (M-En) networks presents significant challenges due to diverse malicious traffic patterns and the evolving nature of cyber threats. Existing AI-based detection systems struggle to adapt to new attack strategies and lack real-time attack detection capabilities with high accuracy and efficiency. This study proposes an online, continuous learning methodology for DDoS detection in M-En networks, enabling continuous model updates and real-time adaptation to emerging threats, including zero-day attacks. First, we develop a unique M-En network dataset by setting up a realistic, real-time simulation using the NS-3 tool, incorporating both victim and bot devices. DDoS attacks with varying packet sizes are simulated using the DDoSim application across IoT and traditional IP-based environments under M-En network criteria. Our approach employs a multi-level framework (MULTI-LF) featuring two machine learning models: a lightweight Model 1 (M1) trained on a selective, critical packet dataset for fast and efficient initial detection, and a more complex, highly accurate Model 2 (M2) trained on extensive data. When M1 exhibits low confidence in its predictions, the decision is escalated to M2 for verification and potential fine-tuning of M1 using insights from M2. If both models demonstrate low confidence, the system flags the incident for human intervention, facilitating model updates with human-verified categories to enhance adaptability to unseen attack patterns. We validate the MULTI-LF through real-world simulations, demonstrating superior classification accuracy of 0.999 and low prediction latency of 0.866 seconds compared to established baselines. Furthermore, we evaluate performance in terms of memory usage (3.632 MB) and CPU utilization (10.05%) in real-time scenarios. Copyright © 2025, The Authors. All rights reserved.

关键词： Multiple access interference

来源：评论

学校读者我要写书评

暂无评论

Structured Process on FL for Big Data Analysis 6

Structured Process on FL for Big Data Analysis

引用

6th International Conference on Mobile Computing and Sustainable informatics, ICMCSI 2025

作者： Govindaram, Anitha Prasath, J.S. Jayasakthi, K. Rajkumar, N. Porkodi, G. Jose Anand, A. Thandalam Chennai602105 India Department of Computer Science and Engineering Sapthagiri NPS University Chikkasandra Karnataka Bengaluru560057 India Department of ECE KCG College of Technology Karapakkam Tamil Nadu Chennai600097 India Department of CSE Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Thiruvalluvar Tamil Nadu Avadi600062 India Department of Computer Science and Business Systems Vel Tech Multi Tech Dr. Rangaranjan Dr.Sakunthala Engineering College Avadi Tamil Nadu Thiruvallur600062 India

ISBN: (纸本)9798331522667

The integration of the Internet of Things (IoT) into various aspects of daily life is accelerated by the proliferation of cognitive capabilities and AI-driven applications. Traditional automated systems rely on centralized data storage and dispensation, which may be feasible in specific application scenarios. However, the increasing flexibility of recent IoT networks and rising concerns over data confidentiality are challenging this approach. Federated Learning (FL), a decentralized collaborative AI technique, allows machine learning to occur directly on external IoT devices without the need for data to be centrally aggregated or shared. This paves the way for a wide variety of intelligent IoT applications to be developed. This article presents a thorough analysis of the along with a detailed analysis of the convergence of these two fields. After that, a full analysis of how FL is currently being utilized in several essential applications related to the IoT is provided. These include smart health, mobility, Unmanned Aerial Vehicles (UAVs), infrastructure improvements, and smart industries. In total, 49 publications from the years 2015-2022 were reviewed here. Finally, discuss future study challenges with issues and exposed questions about data with IoT. © 2025 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Automatic melanoma and non-melanoma skin cancer diagnosis using advanced adaptive fine-tuned convolution neural networks

引用

Discover Oncology 2025年第1期16卷 1-27页

作者： Khan, Muhammad Amir Mazhar, Tehseen Ali, Muhammad Danish Khattak, Umar Farooq Shahzad, Tariq Saeed, Mamoon M. Hamam, Habib School of Computing Sciences College of Computing Informatics and Mathematics Universiti Teknologi MARA Shah Alam 40450 Malaysia School of Computer Science National College of Business Administration and Economics Lahore 54000 Pakistan Department of Computer Science COMSATS University Islamabad Abbottabad Campus Abbottabad 22060 Pakistan School of Information Technology UNITAR International University Kelana Jaya Petaling Jaya 47301 Malaysia Department of Computer Engineering COMSATS University Islamabad Sahiwal Campus Sahiwal 57000 Pakistan Department of Communications and Electronics Engineering Faculty of Engineering University of Modern Sciences (UMS) Yemen Sana’a 00967 Yemen Faculty of Engineering University de Moncton Moncton E1A3E9 NB Canada School of Electrical Engineering Department of Electrical and Electronic Engineering Science University of Johannesburg Johannesburg South Africa Hodmas University College Taleh Area Mogadishu Banadir 521376 Somalia Bridges for Academic Excellence—Spectrum Tunis Centre-Ville 1002 Tunisia Department of Computer Science School Education Department Government of Punjab Layyah 31200 Pakistan

Skin Cancer is an extensive and possibly dangerous disorder that requires early detection for effective treatment. Add specific global statistics on skin cancer prevalence and mortality to emphasize the importance of early detection. Example: “Skin cancer accounts for 1 in 5 diagnosed cancers globally, with melanoma causing over 60,000 deaths annually. Manual skin cancer screening is both time-intensive and expensive. Deep learning (DL) techniques have shown exceptional performance in various applications and have been applied to systematize skin cancer diagnosis. However, training DL models for skin cancer diagnosis is challenging due to limited available data and the risk of overfitting. Traditionally approaches have High computational costs, a lack of interpretability, deal with numerous hyperparameters and spatial variation have always been problems with machine learning (ML) and DL. An innovative method called adaptive learning has been developed to overcome these problems. In this research, we advise an intelligent computer-aided system for automatic skin cancer diagnosis using a two-stage transfer learning approach and Pre-trained Convolutional Neural Networks (CNNs). CNNs are well-suited for learning hierarchical features from images. Annotated skin cancer photographs are utilized to detect ROIs and reset the initial layer of the pre-trained CNN. The lower-level layers learn about the characteristics and patterns of lesions and unaffected areas by fine-tuning the model. To capture high-level, global features specific to skin cancer, we replace the fully connected (FC) layers, responsible for encoding such features, with a new FC layer based on principal component analysis (PCA). This unsupervised technique enables the mining of discriminative features from the skin cancer images, effectively mitigating overfitting concerns and letting the model adjust structural features of skin cancer images, facilitating effective detection of skin cancer features. The sys

关键词： Advanced adaptive fine-tuned Convolution neural networks Melanoma and non-melanoma skin cancer

来源：评论

学校读者我要写书评

暂无评论

AugGPT: Leveraging ChatGPT for Text Data Augmentation

引用

IEEE Transactions on Big Data 2025年第3期11卷 907-918页

作者： Dai, Haixing Liu, Zhengliang Liao, Wenxiong Huang, Xiaoke Cao, Yihan Wu, Zihao Zhao, Lin Xu, Shaochen Zeng, Fang Liu, Wei Liu, Ninghao Li, Sheng Zhu, Dajiang Cai, Hongmin Sun, Lichao Li, Quanzheng Shen, Dinggang Liu, Tianming Li, Xiang University of Georgia School of Computing AthensGA30602 United States South China University of Technology School of Computer Science and Engineering Guangzhou510641 China Lehigh University Department of Computer Science and Engineering BethlehemPA18015 United States Carnegie Mellon University Heinz College of Information Systems and Public Policy PittsburghPA15213 United States Harvard Medical School Department of Radiology Massachusetts General Hospital BostonMA02115 United States Mayo Clinic Department of Radiation Oncology PhoenixAZ85054 United States University of Virginia School of Data Science CharlottesvilleVA22903 United States The University of Texas at Arlington Department of Computer Science and Engineering ArlingtonTX76019 United States ShanghaiTech University School of Biomedical Engineering Shanghai201210 China Shanghai United Imaging Intelligence Company Ltd. Shanghai200230 China Shanghai Clinical Research and Trial Center Shanghai201210 China

Text data augmentation is an effective strategy for overcoming the challenge of limited sample sizes in many natural language processing (NLP) tasks. This challenge is especially prominent in the few-shot learning (FSL) scenario, where the data in the target domain is generally much scarcer and of lowered quality. A natural and widely used strategy to mitigate such challenges is to perform data augmentation to better capture data invariance and increase the sample size. However, current text data augmentation methods either can’t ensure the correct labeling of the generated data (lacking faithfulness), or can’t ensure sufficient diversity in the generated data (lacking compactness), or both. Inspired by the recent success of large language models (LLM), especially the development of ChatGPT, we propose a text data augmentation approach based on ChatGPT (named "AugGPT"). AugGPT rephrases each sentence in the training samples into multiple conceptually similar but semantically different samples. The augmented samples can then be used in downstream model training. Experiment results on multiple few-shot learning text classification tasks show the superior performance of the proposed AugGPT approach over state-of-the-art text data augmentation methods in terms of testing accuracy and distribution of the augmented samples. © 2015 IEEE.

关键词： Zero-shot learning

来源：评论

学校读者我要写书评

暂无评论

GMOCSO: Multi-objective Cat Swarm Optimization Algorithm based on a Grid System

arXiv

引用

arXiv 2025年

作者： Ahmed, Aram M. Hassan, Bryar A. Rashid, Tarik A. Noori, Kaniaw A. Saeed, Soran Ab.M. Ahmed, Omed H. Umar, Shahla U. Computer Science and Engineering Department University of Kurdistan Hewlêr Erbil Iraq Department of Computer Science College of Science Charmo University Chamchamal Sulaimani Iraq Database Technology Department Technical College of Informatics Sulaimani Polytechnic University Sulaimani Iraq Department of Information Technology University of Human Development Sulaymaniyah Iraq Network Department College of Computer Science and Information Technology Kirkuk University Kirkuk Iraq

This paper presents a multi-objective version of the Cat Swarm Optimization Algorithm called the Grid-based Multiobjective Cat Swarm Optimization Algorithm (GMOCSO). Convergence and diversity preservation are the two main goals pursued by modern multi-objective algorithms to yield robust results. To achieve these goals, we first replace the roulette wheel method of the original CSO algorithm with a greedy method. Then, two key concepts from Pareto Archived Evolution Strategy Algorithm (PAES) are adopted: the grid system and double archive strategy. Several test functions and a real-world scenario called the Pressure vessel design problem are used to evaluate the proposed algorithm's performance. In the experiment, the proposed algorithm is compared with other well-known algorithms using different metrics such as Reversed Generational Distance, Spacing metric, and Spread metric. The optimization results show the robustness of the proposed algorithm, and the results are further confirmed using statistical methods and graphs. Finally, conclusions and future directions were presented.. © 2025, CC BY.

关键词： Pressure vessels

来源：评论

学校读者我要写书评

暂无评论

Knowledge Graph-based Retrieval-Augmented Generation for Schema Matching

arXiv

引用

arXiv 2025年

作者： Ma, Chuangtao Chakrabarti, Sriom Khan, Arijit Molnár, Bálint The Department of Computer Science Aalborg University Aalborg Denmark The Department of Information Systems Faculty of Informatics Eötvös Loránd University Budapest Hungary

Traditional similarity-based schema matching methods are incapable of resolving semantic ambiguities and conflicts in domain-specific complex mapping scenarios due to missing commonsense and domain-specific knowledge. The hallucination problem of large language models (LLMs) also makes it challenging for LLM-based schema matching to address the above issues. Therefore, we propose a Knowledge Graph-based Retrieval-Augmented Generation model for Schema Matching, referred to as the KG-RAG4SM. In particular, KG-RAG4SM introduces novel vector-based, graph traversal-based, and query-based graph retrievals, as well as a hybrid approach and ranking schemes that identify the most relevant subgraphs from external large knowledge graphs (KGs). We showcase that KG-based retrieval-augmented LLMs are capable of generating more accurate results for complex matching cases without any re-training. Our experimental results show that KG-RAG4SM outperforms the LLM-based state-of-the-art (SOTA) methods (e.g., Jellyfish-8B) by 35.89% and 30.50% in terms of precision and F1 score on the MIMIC dataset, respectively;KG-RAG4SM with GPT-4o-mini outperforms the pre-trained language model (PLM)-based SOTA methods (e.g., SMAT) by 69.20% and 21.97% in terms of precision and F1 score on the Synthea dataset, respectively. The results also demonstrate that our approach is more efficient in end-to-end schema matching, and scales to retrieve from large KGs. Our case studies on the dataset from the real-world schema matching scenario exhibit that the hallucination problem of LLMs for schema matching is well mitigated by our solution. Copyright © 2025, The Authors. All rights reserved.

关键词： Knowledge graph

来源：评论

学校读者我要写书评

暂无评论

Distill & Contrast: A New Graph Self-Supervised Method With Approximating Nature Data Relationships

引用

IEEE Transactions on Knowledge and Data Engineering 2025年第6期37卷 3284-3297页

作者： He, Dongxiao Zhao, Jitao Guo, Rui Feng, Zhiyong Huo, Cuiying Jin, Di Pedrycz, Witold Zhang, Weixiong Tianjin University College of Intelligence and Computing Tianjin300350 China Department of Measurement and Control Systems Gliwice44-100 Poland University of Alberta Department of Electrical and Computer Engineering EdmontonABT6G 2R3 Canada Istinye University Research Center of Performance and Productivity Analysis Istanbul34010 Turkey Hong Kong Polytechnic University Department of Data Science and Ai Hong Kong

Contrastive Learning (CL) has emerged as a popular self-supervised representation learning paradigm that has been shown in many applications to perform similarly to traditional supervised learning methods. A key component of CL is mining the latent discriminative relationships between positive and negative samples and using them as self-supervised labels. We argue that this discriminative contrastive task is, in essence, similar to a classification task, and the "either positive or negative"hard label sampling strategies are arbitrary. To solve this problem, we explore ideas from data distillation, which considers probabilistic logit vectors as soft labels to transfer model knowledge. We attempt to abandon the classical hard sampling labels in CL and instead explore self-supervised soft labels. We adopt soft sampling labels that are extracted, without supervision, from the inherent relationships in data pairs to retain more information. We propose a new self-supervised graph learning method, Distill and Contrast (D&C), for learning representations that closely approximate natural data relationships. D&C extracts node similarities from the features and structures to derive soft sampling labels, which also eliminate noise in the data to increase robustness. Extensive experimental results on real-world datasets demonstrate the effectiveness of the proposed method. © 2025 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

A Design of Hybrid Model and Bayesian Neural Networks for Smart Grid Stability Prediction 1

A Design of Hybrid Model and Bayesian Neural Networks for Sm...

引用

1st IEEE International Conference on Smart and Sustainable Developments in Electrical Engineering, SSDEE 2025

作者： Hole, Shreyas Rajendra Kolluru, Vinothkumar Salotagi, Shreekant Challagundla, Yagnesh Mungara, Sudeep Sriramkumar, R. Dayananda Sagar University AI & ML Bangalore India Stevens Institute of Technology Department of Data Science HobokenNJ07030 United States Dayananda Sagar University Computer Science & Engineering Bagalore India Herbert Wertheim College of Engineering University of Florida Department of Computer Science GainesvilleFL32611 United States Stevens Institute of Technology Masters in Information Systems HobokenNJ07030 United States

ISBN: (纸本)9798331542108

The latest evolution in power systems, is 'smart grids'that offers real time monitoring, control features as well as effective management of renewable energy sources. Nonetheless, with the increase in the system complexity, the chances for insta- bility also increase and thus it is crucial to forecast the stability accurately for the safety of the grid. Weather it is in terms of Gradient Boosting Tree Algorithm or any other traditional algorithm, these machine learning models are successful in achieving the objectives of the task but they lack a forthcoming metric to assist in high stake situations, uncertainty quantification. The paper is devoted to the analysis of the reliability of smart grid systems based on Bayesian Neural Networks (BNN) taking into account both the reliability of predictive modeling and forecasting uncertainty. The smart grid stability augmented dataset was used to benchmark the efficacy of using BNNs or BNNs combined hybrid models, one with XGBoost, for feature extraction and the other with LightGBM. The present study shows that the BNN used with the above mentioned smart grid system achieved an accuracy of 92% with precision of 0.89, recall of 0.91 and F1 of 0.90. These metrics are appending with the help of the additional features of BNN where they are capable of addressing the key questions around uncertainty and thus the metrics confirming the predictions are quite strong in terms of real-time grid stability evaluation. This article encapsulates a detailed study of BNN architecture, the role of uncertainty in machine learning models and the natural application of such models in the design of dependable and fault-tolerant smart grids. © 2025 IEEE.

关键词： Smart power grids

来源：评论

学校读者我要写书评

暂无评论

Adaptive iterated local search algorithm for dynamic patient admission scheduling problems

引用

Soft Computing 2025年 1-11页

作者： Turky, Ayad Sabar, Nasser R. Song, Andy Hussain, Abir Liatsis, Panos Department of Computer Science College of Computing and Informatics University of Sharjah Sharjah United Arab Emirates Department of Computer Science and Information Technology La Trobe University Melbourne 3086 VIC Australia School of Computing Technologies RMIT University Melbourne Australia Department of Electrical Engineering University of Sharjah Sharjah United Arab Emirates Department of Electrical Engineering and Computer Science Khalifa University Abu Dhabi United Arab Emirates

Healthcare resource management is essential for ensuring the quality of patient care. However, it can be a complex and costly task. This work addresses the patient admission scheduling (PAS) problem, a complex aspect of healthcare resource management. PAS aims to allocate patients to hospital beds within a planning horizon, subject to a variety of healthcare constraints. The goal is to maximize management efficiency and patient comfort to improve medical treatment. In this work, we consider a practical variant of PAS known as dynamic PAS (DPAS). DPAS considers several factors and constraints, such as the daily registration of new patients, urgent patients, uncertainties in stay lengths, operating theatre resources, and the surgery scheduling process. An effective and efficient adaptive iterated local search (AILS) algorithm is proposed to solve DPAS. To enable the search to explore the search space efficiently, the proposed AILS adaptively integrates a number of components. Two adaptive perturbation strategies are devised to locate unexplored areas in the search space. To exploit the newly discovered areas effectively, we propose an adaptive local search mechanism as an intensification strategy to find a high-quality solution. The proposed AILS algorithm is compared to benchmark problems used by existing algorithms. The experimental results demonstrate the effectiveness and efficiency of the proposed approach. Specifically, out of 30 tested instances, AILS obtains 17 of the best-known results using less computational time. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Dynamic patient admission scheduling Healthcare resource management Iterated local search algorithm Patient bed assignment

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：