检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Ning, Kaiwen Chen, Jiachi Zhang, Jingwen Li, Wei Wang, Zexu Feng, Yuming Zhang, Weizhe Zheng, Zibin The School of Software Engineering Sun Yat-Sen University Peng Cheng Laboratory China The School of Software Engineering Sun Yat-Sen University China The School of Computer Science and Technology Harbin Institute of Technology China Peng Cheng Laboratory China The Department of New Networks Peng Cheng Laboratory China

Artificial intelligence (AI) agents are systems capable of perceiving their environment, autonomously planning and executing tasks. Recent advancements in Large Language Models (LLMs) have introduced a transformative paradigm for AI agents, enabling them to interact with external resources and tools through prompt techniques. This advancement has significantly extended the capabilities of LLMs, positioning LLM-based AI Agents as an important research area. In such agents, the workflow integrates developer-written code, which manages framework construction and logic control, with LLM-generated natural language that enhances dynamic decision-making and interaction. However, discrepancies between developer-implemented logic and the dynamically generated content of LLMs in terms of behavior and expected outcomes can lead to defects, such as tool invocation failures and task execution errors. These issues introduce specific risks, leading to various defects in LLM-based AI Agents, including service interruptions and incorrect output. Despite the importance of these issues, there is a lack of systematic work that focuses on analyzing LLM-based AI Agents to uncover defects in their code. To address this gap, we present the first study focused on identifying and detecting defects in LLM Agents. We collected and analyzed 6, 854 relevant posts from StackOverflow to define and classify eight types of agent defects. For each defect type, we provided detailed descriptions and illustrated with an example. Then, we designed a static analysis tool, named Agentable, to detect the defined defects. Agentable leverages Code Property Graphs (CPGs) and LLMs to analyze Agent workflows by efficiently identifying specific code patterns and analyzing natural language descriptions. To evaluate Agentable, we constructed two datasets: AgentSet, which consists of 84 real-world Agent projects, and AgentTest, which contains 78 Agent projects specifically designed to include various types of defect

关键词： Autonomous agents

来源：评论

学校读者我要写书评

暂无评论

Development of a New Opto-Electrochemical Cell for Sensing Applications

Development of a New Opto-Electrochemical Cell for Sensing A...

引用

IEEE Sensors Applications Symposium, SAS

作者： João P. Mendes Luís C. C. Coelho José A. Ribeiro Centre for Applied Photonics INESCTEC - Institute for Systems and Computer Engineering Technology and Science Porto Portugal Department of Chemistry and Biochemistry CIQUP/IMS Faculty of Sciences of University of Porto Porto Portugal

ISBN: (数字)9798350369250

ISBN: (纸本)9798350369267

New systems with innovative design to perform measurements combining electrochemistry and surface plasmon resonance (ESPR) are currently a need to overcome the limitations of existent market solutions and expand the research possibilities of this technology. The main goal of this work was to develop a new cell to increase ESPR practical applications in several fields. To do so, a homemade SPR cell, fabricated by 3D-printing technology, was adapted for this purpose by incorporating the conventional 3-electrodes to perform the electrochemical experiments. The developed cell was fully compatible with commercial SPR substrates. After optimization of the homemade ESPR setup to perform the combined electrochemical and SPR measurements, two main applications were explored in this work. The first was the use of ESPR technology as straightforward tool to simultaneously investigate the electrical and optical properties of conducing/non-conducting polymers electrosynthetized on the SPR platforms. The conducting polymer poly(thionine) was used in this work for proof-of-concept. The second application envisaged the use of ESPR approach for simple electrodeposition ofmaterials with enhanced plasmonic properties for sensitivity enhancement of SPR biosensors. For validation of the concept, graphene oxide (GO) was electrochemically reduced on gold substrates aiming to evaluate the plasmonic properties of graphene-modified sensing surfaces.

关键词： Gold Biomedical optical imaging Sensitivity Microprocessors computer architecture Stability analysis Plastics

来源：评论

学校读者我要写书评

暂无评论

Enhancing Sovereign Allocation of Resources in Cloud Milieus Using a Causal Dilated Geometric Algebra Approach for Dynamic Scalability

Enhancing Sovereign Allocation of Resources in Cloud Milieus...

引用

Machine Learning and Autonomous systems (ICMLAS), International Conference on

作者： Chandrasekhar Rao Katru Sathish Srinivasan Mohan Kumar T G Sunita Yadwad G. Satish Ramya Maranan Lead Software Engineer Indian Land SC USA Oracle Cloud Infrastructure Division Machine Learning & AI Development San Francisco Bay Area California USA Department of Information Science and Engineering Nitte Meenakshi Institute of Technology Bengaluru Karnataka Computer science and Engineering Faculty of Engineering and Technology Parul institute of Engineering and Technology Parul University Post Limda Gujarat Department of Electrical and Electronics Engineering Sree Dattha Institute of Engineering and Science Sheriguda Telangana Department of Research and Innovation Saveetha School of Engineering SIMATS Chennai Tamil Nadu

ISBN: (数字)9798331505745

ISBN: (纸本)9798331505752

Cloud environments utilize availability and usage of resources as crucial components in order to produce scalability, reliability and low cost. Such resource capabilities require autonomous systems that can handle these resources dynamically in spite of fluctuating workloads and service delivery requirements. Static models of resource management suffer from scalability and inability to react on the fly and accurately predict the demand thus causing Service Level Agreement (SLA) breaches and affecting the Quality of Service (QoS) as well as energy consumption profile. Advanced and elaborative cloud workloads require an intelligent system that can analyse the temporal and spatial patterns without impacting the overall system's performance significantly. Inspired by these challenges, this research proposes a Causal Dilated Geometric Algebra Convolutional Transformer Network (CATNet) along with the Prairie-Dog Optimizer (P-DogO), referred to as CAT-DogNet to learn and predict the dynamic scalability. CATNet integrates Causal Dilated Convolutional Networks in time domain for temporal structure capturing and Geometric Algebra Transformer for spatial resource coding. P-DogO adjusts weight parameters and losses for resource allocation, enhancing them depending on a work environment. A set of comprehensive assessments demonstrate a 0.85% SLA violation rate, average response time of less than 10ms, and overall 12.3% less energy consumption than baseline methods. The system had favourable predictability rates over 99.4% for the resource demand estimate. CAT-DogN et presents a sound and agnostic architecture for autonomous resource allocation in cloud environment, which indeed extend the state of art by considering flexibility, power consumption and QoS compliance.

关键词： Cloud computing Energy consumption Algebra Scalability Quality of service Transformers Real-time systems Resource management Time factors Service level agreements

来源：评论

学校读者我要写书评

暂无评论

Facial Skincare Product Suggestion with Product Popularity and Post Recommendation Care

Facial Skincare Product Suggestion with Product Popularity a...

引用

2023 4th International Informatics and software engineering Conference (IISEC)

作者： H. M. W. T. C. B. Samarakoon A. U Ovitigala G. A. T. T. Wijesinghe K. M. L. P. Weerasinghe G. K. B. V. Karunathilake Ishara Weerathunga Department of Computer Science and Software Engineering Sri Lanka Institute of Information Technology Malabe Sri Lanka Department of Information Technology Sri Lanka Institute of Information Technology Malabe Sri Lanka

Skin health significantly affects a person's overall wellbeing, which includes both physical and emotional aspects. The objective of this study is to develop an integrated system for recommending skincare products that integrates image processing, machine learning, sentiment analysis, and user feedback analysis to deliver individualized skincare advice and boost user satisfaction. Utilizing deep learning algorithms and face cues, the system uses user image analysis to detect important skin features, skin type, acne severity, and dark spots with extraordinary accuracy. Using sentiment analysis and clustering, the program identifies popular skincare products and tailors’ recommendations to user preferences. Recommendations are further personalized by collaborative filtering. By using image analysis to identify allergies and other negative effects, user safety is increased. The technique prioritizes data safety and includes model training, sentiment analysis, image processing, and user input analysis. Initial findings demonstrate the excellent accuracy of skin issue detection and severity prediction. By allowing users from a variety of user demographics to make informed skincare decisions, this complete framework has the potential to change the skincare industry.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Development of a Prediction System for Criminality and Suicidal in Sri Lanka

Development of a Prediction System for Criminality and Suici...

引用

Advancements in Computing (ICAC), International Conference on

作者： D.P.D. Ranasinghe R.D.T.S. Fernando D.G. Girambe D.M.H.K.D. Arambegama Bhagyanie Chathurika Dulani Jayasinghe Department of Computer Science and Software Engineering Sri Lanka Institute of Information Technology Malabe Sri Lanka Department of Information Technology Sri Lanka Institute of Information Technology Malabe Sri Lanka

This research confronts the growing challenges of rising suicide rates and crime in Sri Lanka through a dual approach, harnessing advanced machine learning and time series analysis. The innovative methodology, merging “Feature Importance Weighting and Ensemble of Diverse Trees with the Random Forest Regressor,” achieves impressive accuracy of 93% for factors and 94% for methods. Additionally, the accuracy rates of 94% for education and 93% for occupationbased predictions enhance this comprehensive approach, offering a profound understanding of suicidal incidents and enabling precise prevention strategies. Simultaneously this study employs an ensemble of AutoRegressive Integrated Moving Average (ARIMA) models for crime prediction. This approach consistently attains high accuracy, often surpassing 90%, such as 96% and 97%. Utilizing multiple models enhances predictive accuracy and hotspot identification, granting law enforcement and policymakers flexibility to address enduring yearly hotspots and evolving dynamic hotspots. This research bridges the comprehension of suicide and crime dynamics in Sri Lanka, providing invaluable insights for evidence-based policies and targeted interventions.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A Kernel Propagation-Based Graph Convolutional Network Imbalanced Node Classification Model on Graph Data 19

A Kernel Propagation-Based Graph Convolutional Network Imbal...

引用

19th IEEE International Conference on Networking, Sensing and Control, ICNSC 2022

作者： Bai, Xiang-En An, Jing Yu, Zi-Bo Bao, Han-Qiu Wang, Ke-Fan School of Computer Science and Information Engineering Shanghai Institute of Technology Shanghai China College of Applied Mathematics Xi'an Jiaotong-Liverpool University Suzhou China College of Electronic and Information Engineering Tongji University Shanghai China Shanghai Research Institute for Intelligent Autonomous Systems Tongji University Shanghai China

ISBN: (纸本)9781665472432

As an important research topic in graph learning, the performance of node classification has been improved with the development of some new methods, among which graph neural networks (GNNs) have achieved state-of-the-art node classification performance. However, the existing GNN-based methods mainly address the classification problem with balanced distribution of node samples. However, many real application scenarios of graph data usually have a highly skewed class distribution, i.e., the majority classes occupy most of the samples while the minority classes contain only a few samples. When the nodes exhibit an imbalanced class distribution, existing GNN-based methods favor the majority class and under-represent the minority class. Therefore, we propose a novel Kernel Propagation-based model for Imbalanced Node Classification in Graph Convolutional Networks (KINC-GCN). First, we introduce a kernel propagation method as a preprocessing step to exploit higher-order structural features. The node features are enhanced by concatenating the higher-order structural feature matrix with the node feature matrix. Node embeddings are obtained from the enhanced feature and adjacency matrices by a two-layer GCN, and then a self-optimizing cluster analysis and graph reconstruction module are introduced. The self-optimizing cluster analysis module performs cluster analysis on the node embeddings to enhance the representativeness of the node embeddings. The graph reconstruction module uses an inner product decoder to reconstruct the graph structure and minimize the differences between the reconstructed graph and the original graph. The effectiveness of KINC-GCN in node classification is demonstrated by experiments on three real-world imbalanced graph datasets. © 2022 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

A novel approach incorporating feature extraction followed by YOLOv7 for quality enhancement of mangoes in Bangladesh

A novel approach incorporating feature extraction followed b...

引用

International Conference on Computing and Networking Technology (ICCNT)

作者： Debashis Kundu Md. Mushfiqur Rahaman Md. Nazmus Sakib Md. Faruk Abdullah Al Sohan Jannatul Fardaush Mithu Afzal Hossain Nishita Aktar Dept. Management Information Systems American International University- Bangladesh Dhaka Bangladesh Dept. of Computer Science and Engineering Daffodil Institute of IT Dhaka Bangladesh Dept. of Computer Science and Engineering World University of Bangladesh Dhaka Bangladesh Dept. of Computer Science American International University- Bangladesh Dhaka Bangladesh Dept. of Electrical and Computer Engineering North South University Dhaka Bangladesh Dept. of Computer Science and Engineering Army Institute of Business Administration (AIBA) Savar Dhaka Bangladesh

Qualitative approach for automated grading and quality assessment of fruits, machine learning techniques are crucial in agricultural applications. Automation enhances a nation’s agricultural quality, production, and economic prosperity. Fruit quality grading, particularly the surface fault identification of a fruit, is a crucial indicator in the export market. This is particularly important for mangoes, which are quite well-liked inBangladesh. On the other hand, the physical grading of mangoes is a procedure that is labor-intensive, prone to error, and very subjective. In this paper, we proposed a YOLOv7 integrated Discrete wave transformation computer vision system. The proposed model includes support vector machine (SVM) and decision tree for the classification of high-quality mangoes. The results of the experiments show that the proposed solution obtained 96.25% accuracy when the system was trained and tested using a publicly accessible mango database.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Generalized Correspondence Matching via Flexible Hierarchical Refinement and Patch Descriptor Distillation

Generalized Correspondence Matching via Flexible Hierarchica...

引用

IEEE International Conference on Robotics and Automation (ICRA)

作者： Yu Han Ziwei Long Yanting Zhang Jin Wu Zhijun Fang Rui Fan School of Computer Science and Technology Donghua University Shanghai P. R. China Shanghai Research Institute for Intelligent Autonomous Systems the State Key Laboratory of Intelligent Autonomous Systems and Frontiers Science Center for Intelligent Autonomous Systems Machine Intelligence & Autonomous Systems (MIAS) Group the College of Electronics & Information Engineering Tongji University Shanghai P. R. China Department of Electronics and Computer Engineering The Hong Kong University of Science and Technology P. R. China

ISBN: (数字)9798350384574

ISBN: (纸本)9798350384581

Correspondence matching plays a crucial role in numerous robotics applications. In comparison to conventional hand-crafted methods and recent data-driven approaches, there is significant interest in plug-and-play algorithms that make full use of pre-trained backbone networks for multi-scale feature extraction and leverage hierarchical refinement strategies to generate matched correspondences. The primary focus of this paper is to address the limitations of deep feature matching (DFM), a state-of-the-art (SoTA) plug-and-play correspondence matching approach. First, we eliminate the pre-defined threshold employed in the hierarchical refinement process of DFM by leveraging a more flexible nearest neighbor search strategy, thereby preventing the exclusion of repetitive yet valid matches during the early stages. Our second technical contribution is the integration of a patch descriptor, which extends the applicability of DFM to accommodate a wide range of backbone networks pre-trained across diverse computer vision tasks, including image classification, semantic segmentation, and stereo matching. Taking into account the practical applicability of our method in real-world robotics applications, we also propose a novel patch descriptor distillation strategy to further reduce the computational complexity of correspondence matching. Extensive experiments conducted on three public datasets demonstrate the superior performance of our proposed method. Specifically, it achieves an overall performance in terms of mean matching accuracy of 0.68, 0.92, and 0.95 with respect to the tolerances of 1, 3, and 5 pixels, respectively, on the HPatches dataset, outperforming all other SoTA algorithms. Our source code, demo video, and supplement are publicly available at ***/GCM.

关键词： Performance evaluation Accuracy Semantic segmentation Source coding Image matching Supervised learning Pose estimation

来源：评论

学校读者我要写书评

暂无评论

A Comprehensive Evaluation of Q-Learning Based Automatic Web GUI Testing

A Comprehensive Evaluation of Q-Learning Based Automatic Web...

引用

International Conference on Dependable systems and Their Applications (DSA)

作者： Yujia Fan Siyi Wang Sinan Wang Yepang Liu Guoyao Wen Qi Rong Department of Computer Science and Engineering Southern University of Science and Technology Shenzhen China Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology Shenzhen China Huawei Inc China

Recently, reinforcement learning (RL) based automatic Web GUI testing techniques are gaining popularity in both academia and industry as they can enable more intelligent exploration of web applications’ states. However, the existing RL-based techniques often incorporate special features, such as DFA-guided state recovery or contextual input data generation, making the effectiveness of RL itself unclear. Moreover, while these techniques mostly employ Q-learning (QL), a model-free RL method, they were evaluated with different experimental settings, which could lead to unfair comparisons. Motivated by the two observations, we propose a generic QL-based automatic Web GUI testing framework, and conduct the first systematic evaluation, which considers four QL specific configurations, on two open-source benchmark web applications and one industrial portal website. Based on the experimental results, we discuss several findings regarding the effectiveness of QL-based automatic GUI testing. We believe that our findings can provide useful guidance to industrial practitioners and shed light on future research on leveraging RL to improve automatic Web GUI testing.

关键词：

来源：评论

学校读者我要写书评

暂无评论

PointGAN: A Catalyst for Enhanced Vulnerable Road User Detection in Autonomous Navigation

PointGAN: A Catalyst for Enhanced Vulnerable Road User Detec...

引用

IEEE Conference on Vehicular Technology (VTC)

作者： Gunasekaran Raja Hosam Alhakami Sahaya Beni Prathiba J eyadev N eedhidevan Priyadarshni Vasudevan Rupali Subramanian Kapal Dev Department of Computer Technology NGNLab Anna University Chennai India College of Computing Umm AI-Qura University Makkah Saudi Arabia Centre for Cyber Physical Systems School of Computer Science and Engineering Vellore Institute of Technology Chennai India Department of Computer Science Munster Technological University Ireland

ISBN: (数字)9798350387414

ISBN: (纸本)9798350387421

In autonomous vehicle navigation, the effectiveness and robustness of object detection models are directly influenced by the quality, quantity, and diversity of data. Detecting Vul-nerable Road Users (VRU) poses significant challenges due to their constant motion and dynamic behaviors. Utilizing real-time datasets like KITTI for training VRU detection models may not be optimal, as they lack coverage of adversarial situations and undervalue objects like pedestrians. Relying solely on such datasets for model training can lead to catastrophic real-world results. To address these issues, we present PointGAN, a framework supporting 3D object detection models that leverage conditional Generative Adversarial Networks (cGAN) to enhance dataset diversity specifically for the pedestrian class. PointGAN employs generative neural networks trained through multiple iterations to generate realistic point cloud data, guided by feedback from the discriminator until it closely mirrors real-world data. This strategy intends to significantly improve the overall performance of 3D object detection models by skillfully detecting pedestrians in challenging and diverse scenarios. The generative model achieves a Minimum Matching Distance - Earth Mover's Distance (MMD-EMD) score of 0.025, outperforming the existing state-of-the-art models trained under different categories.

关键词： Training Point cloud compression Solid modeling Pedestrians Three-dimensional displays Roads Object detection

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：