检索结果-内蒙古大学图书馆

41st International Conference on Machine Learning, ICML 2024

作者： Zhang, Wanpeng Li, Yilin Yang, Boyu Lu, Zongqing School of Computer Science Peking University China Center for Statistical Science Peking University China School of Data Science Fudan University China Beijing Academy of Artificial Intelligence China

In real-world scenarios, the application of reinforcement learning is significantly challenged by complex *** existing methods attempt to model changes in the environment explicitly, often requiring impractical prior knowledge of *** this paper, we propose a new perspective, positing that non-stationarity can propagate and accumulate through complex causal relationships during state transitions, thereby compounding its sophistication and affecting policy *** believe that this challenge can be more effectively addressed by implicitly tracing the causal origin of *** this end, we introduce the Causal-Origin REPresentation (COREP) *** primarily employs a guided updating mechanism to learn a stable graph representation for the state, termed as causal-origin *** leveraging this representation, the learned policy exhibits impressive resilience to *** supplement our approach with a theoretical analysis grounded in the causal interpretation for non-stationary reinforcement learning, advocating for the validity of the causal-origin *** results further demonstrate the superior performance of COREP over existing methods in tackling non-stationarity *** code is available at https://***/PKURL/COREP. Copyright 2024 by the author(s)

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Improving chinese cross-modal retrieval with multi-modal transportation data 7

Improving chinese cross-modal retrieval with multi-modal tra...

引用

2024 7th International Symposium on Big data and Applied Statistics, ISBDAS 2024

作者： He, Yanzhong Zhang, Yanjiao Zhu, Lin School of Computer Science and Artificial Intelligence Wuhan University of Technology Wuhan China School of Computer Wuhan Qingchuan University Wuhan China

As societal development progresses and individual travel needs evolve, the demand for acquiring multi-modal transportation status information has steadily increased. The present domain of transportation status encompasses a wealth of multi-modal information, including vehicle trajectory data, traffic condition visual imagery, and textual information. Acquiring multi-modal transportation status information facilitates a rapid understanding of the prevailing traffic conditions at a given location. In this study, we investigate multi-modal transportation status data encompassing trajectories, textual descriptions, and visual imagery. We establish a multi-modal dataset comprising circuit, text, and image data and enhance the state-of-the-art cross-modal retrieval model within the Chinese domain. We utilize hash codes to compute the similarity between images and textual descriptions. During model evaluation using our collected dataset, the Chinese-CLIP-Hash model (our model) demonstrates performance improvements over the Chinese-CLIP model. Specifically, it exhibits a 1.0% improvement in R@10 for the Text-to-Image task and a 0.4% improvement in R@5 for the Image-to-Text task, along with a 0.2% enhancement in R@10. Furthermore, during the cross-modal retrieval process, it achieves a notable speed increase, exhibiting an 11% improvement in speed compared to the Chinese-CLIP model when handling queries from a set of 400 textual descriptions. © Published under licence by IOP Publishing Ltd.

关键词： Network security

来源：评论

学校读者我要写书评

暂无评论

LegoDroid: flexible Android app decomposition and instant installation

引用

science China(Information sciences) 2023年第4期66卷 155-173页

作者： Yi LIU Yun MA Xusheng XIAO Tao XIE Xuanzhe LIU School of Computer Science Peking University Key Laboratory of High Confidence Software Technologies(Peking University) Ministry of Education Institute for Artificial Intelligence Peking University Department of Computer and Data Sciences Case Western Reserve University

Current mobile applications(apps) have become increasingly complicated with increasing features that are represented on the graphical user interface associated with application programming interfaces(APIs) to access backend functionality and data. Meanwhile, apps suffer from the “software bloat” in volume. Some app features may be redundant, with respect to those features(from other apps) that the users already desirably and frequently use. However, the current app release model forces users to download and install a full-size installation package rather than optionally choosing only their desired features. Large-size apps can not only increase the local resource consumption, such as CPU, memory, and energy, but also inevitably compromise the user experience, such as the slow load time in the app. In this article, we first conduct an empirical study to characterize the app feature usage when users interact with Android apps,and surprisingly find that users access only a very small subset of app features. Based on these findings,we design a new approach named Lego Droid, which automatically decomposes an Android app for flexible loading and installation, while preserving the expected functionality with a fast and instant app load. With such a method, a slimmer bundle will be downloaded and host the target APIs inside the original app to satisfy users' requirements. We implement a system for Lego Droid and evaluate it with 1000 real-world Android apps. Compared to the original full-size apps, apps optimized by Lego Droid can substantially improve the load time by reducing the base bundle and feature bundles by 13.06% and 10.93%, respectively,along with the app-package installation size by 44.17%. In addition, we also demonstrate that Lego Droid is quite practical with evolving versions, as it can produce substantial reusable code to alleviate the developers' efforts when releasing new app versions.

关键词： performance software bloat instant installation mobile applications program analysis

来源：评论

学校读者我要写书评

暂无评论

The Classification of Software Requirements using Large Language Models: Emerging Results and Future Directions 3

The Classification of Software Requirements using Large Lang...

引用

3rd International Conference on Cloud Computing, Big data Application and Software Engineering, CBASE 2024

作者： Zhang, Yizhuo Li, Yanhui School of Computer Science and Artificial Intelligence Wuhan University of Technology Wuhan China School of Computer Wuhan Qingchuan University Wuhan China

ISBN: (纸本)9798331506582

In recent years, large language models (LLMs) have demonstrated exceptional capabilities across various tasks. However, the effectiveness of LLMs in software requirement classification has yet to be thoroughly explored. This paper aims to bridge this gap by evaluating the performance of multiple LLMs (including GPT-3.5, GPT-4, GPT-4o, HuggingChat, Gemini, and DeepSeek) under different task prompts, focusing specifically on the binary classification of functional requirements (FR) versus non-functional requirements (NFR), as well as the multi-class classification of non-functional requirements. Experimental results show that HuggingChat achieved the highest accuracy and F1 score in the binary classification task, outperforming other models;meanwhile, GPT-4 performed relatively well in the multi-class classification task. Through the analysis of these results, this paper provides important references for future research and applications in software requirement classification and explores potential directions for LLM-based software requirement classification. © 2024 IEEE.

关键词： Application programs

来源：评论

学校读者我要写书评

暂无评论

Cat Swarm Algorithm Generated Based on Genetic Programming Framework Applied in Digital Watermarking

引用

computers, Materials & Continua 2025年第5期83卷 3135-3163页

作者： Shu-Chuan Chu Libin Fu Jeng-Shyang Pan Xingsi Xue Min Liu School of Artificial Intelligence Nanjing University of Information Science and TechnologyNanjing210000China College of Computer Science and Engineering Shandong University of Science and TechnologyQingdao266000China Department of Information Management Chaoyang University of TechnologyTaichung40601TaiwanChina School of Information Science and Engineering Fujian University of TechnologyFuzhou350000China Key Laboratory of Hydrometeorological Disaster Mechanism and Warning of Ministry of Water Resource Nanjing University of Information Science and TechnologyNanjing210000China

Evolutionary algorithms have been extensively utilized in practical ***,manually designed population updating formulas are inherently prone to the subjective influence of the *** programming(GP),characterized by its tree-based solution structure,is a widely adopted technique for optimizing the structure of mathematical models tailored to real-world *** paper introduces a GP-based framework(GPEAs)for the autonomous generation of update formulas,aiming to reduce human *** modifications to tree-based GP have been instigated,encompassing adjustments to its initialization process and fundamental update operations such as crossover and mutation within the *** designing suitable function sets and terminal sets tailored to the selected evolutionary algorithm,and ultimately derive an improved update *** Cat Swarm Optimization Algorithm(CSO)is chosen as a case study,and the GP-EAs is employed to regenerate the speed update formulas of the *** validate the feasibility of the GP-EAs,the comprehensive performance of the enhanced algorithm(GP-CSO)was evaluated on the CEC2017 benchmark ***,GP-CSO is applied to deduce suitable embedding factors,thereby improving the robustness of the digital watermarking *** experimental results indicate that the update formulas generated through training with GP-EAs possess excellent performance scalability and practical application proficiency.

关键词： Cat swarm algorithm genetic programming digital watermarking update mode mode generation framework

来源：评论

学校读者我要写书评

暂无评论

Text Reasoning Chain Extraction for Multi-Hop Question Answering

引用

Tsinghua science and technology 2024年第4期29卷 959-970页

作者： Pengming Wang Zijiang Zhu Qing Chen Weihuang Dai School of Data Science and Artificial Intelligence Wenzhou University of TechnologyWenzhou 325035China School of Computer Science and also with Institute of Intelligent Information ProcessingSouth China Business CollegeGuangdong University of Foreign StudiesGuangzhou 510545China School of Psychology Jiangxi Normal UniversityNanchang 330022China also with School of Science East China Jiaotong UniversityNanchang 330013China Institute of Intelligent Information Processing South China Business CollegeGuangdong University of Foreign StudiesGuangzhou 510545China

With the advent of the information age, it will be more troublesome to search for a lot of relevant knowledge to find the information you need. Text reasoning is a very basic and important part of multi-hop question and answer tasks. This paper aims to study the integrity, uniformity, and speed of computational intelligence inference data capabilities. That is why multi-hop reasoning came into being, but it is still in its infancy, that is, it is far from enough to conduct multi-hop question and answer questions, such as search breadth, process complexity, response speed, comprehensiveness of information, etc. This paper makes a text comparison between traditional information retrieval and computational intelligence through corpus relevancy and other computing methods. The study finds that in the face of multi-hop question and answer reasoning, the reasoning data that traditional retrieval methods lagged behind in intelligence are about 35% worse. It shows that computational intelligence would be more complete, unified, and faster than traditional retrieval methods. This paper also introduces the relevant points of text reasoning and describes the process of the multi-hop question answering system, as well as the subsequent discussions and expectations.

关键词： intelligent computing multi-hop quiz text reasoning document retrieval text complex network

来源：评论

学校读者我要写书评

暂无评论

ML-Based College Predictor Telegram Bot Indian Engineering Colleges 1

ML-Based College Predictor Telegram Bot Indian Engineering C...

引用

1st International Conference on AIML-Applications for Engineering and technology, ICAET 2025

作者： More, Amruta Patil, Satchal Savale, Aditya Satle, Prafull Tilwane, Krishna Shinde, Sanket Ajeenkya DY Patil School of Engineering Department of Artificial Intelligence and Data Science Pune India

ISBN: (纸本)9798350355611

This Research aimed at the challenge of finding college for students, but this is complex task as students are unable to find best institutes based on entrance exam scores in India. To bounce on that, we focus on the development of a College Predictor Bot, which uses some key factors notably JEE and CET scores and predicts admission to different Indian colleges. You are trained on years of admissions data from hundreds of colleges, with students of all shapes and sizes and performance levels. In terms of methodology used in the study, a combination of random forest and decision tree models were used for analysing the exam scores and make findings. The ML-model is subject to a lot of statistics and testing, and the analysis results in highly valuable co-relations between performance on entrance exams and admitting results. Our results show that College Predictor Bot accurately predicts admissions, therefore helping students know what colleges are suitable. The larger implications [of this tool] are that this might make the college selection process easier and provide a slightly more transparent and informed way of getting deserved seat in the Eductaion system of Bharat. © 2025 IEEE.

关键词： Students

来源：评论

学校读者我要写书评

暂无评论

Online Network Intrusion Detection Based on Self-Encoder Integration 24

Online Network Intrusion Detection Based on Self-Encoder Int...

引用

3rd International Conference on Signal Processing, computer Networks and Communications, SPCNC 2024

作者： Sun, Dongxu Li, Peng Pei, Lina Wei, Chuhang Co. Ltd. Hebei Shijiazhuang China School of School of Artificial Intelligence and Data Science Hebei University of Technology Tianjin China

ISBN: (纸本)9798400710834

With the development of information technology, network security problems have intensified, and the performance enhancement of network intrusion detection systems has attracted much attention In this paper, we propose an online system based on self-encoder integration that utilizes deep learning to improve detection accuracy and robustness Neural networks are popular for their ability to learn complex patterns, but training resources are in high demand and existing supervised schemes require expert labeling and regular updates. Therefore, this paper introduces Kitsune, which uses unsupervised, efficient online self-encoder integration to distinguish between normal and abnormal traffic and is optimized to improve detection efficiency In this paper, we experimentally validate that the optimized Kitsune achieves improvements in accuracy, recall, false-positive rate, false-negative rate. © 2024 Copyright held by the owner/author(s).

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Extended Multi-scale Feature Fusion and Balanced Generative Adversarial Network for Image Inpainting under Limited data

Extended Multi-scale Feature Fusion and Balanced Generative ...

引用

2024 Workshop on Electronics Communication Engineering, WECE 2024

作者： Xiao, Ting Sun, Minqian Chen, Haoyuan Li, Jiaqi Qiu, Yixing Wang, Zhe Lu, Dan Department of Computer Science and Engineering East China University of Science and Technology China School of Artificial Intelligence OPtics and ElectroNics Northwestern Polytechnical University China School of Computer Science and Technology Harbin Engineering University China

ISBN: (数字)9781510689145

ISBN: (纸本)9781510689138

Image inpainting has made remarkable progress in recent years due to the availability of sufficient training data. However, when the training data is insufficient, i.e., image inpainting in few-shot regimes, the performance of existing image inpainting methods degrades drastically. Two challenges impede image inpainting under few-shot regimes, adequate information and unstable training. To address the above issues, this paper proposes a new method for image inpainting under limited data. Specifically, an Extended Multi-scale Feature Fusion (EMFF) module is proposed to capture representative contextual information by progressively reducing the number of channels for features with increasing receptive fields, thereby utilizing effective information. Additionally, we leverage a multi-scale discriminator whose cross-layer features are mixed for supplementation to formulate a Balanced Generative Adversarial Network (BGAN) to stabilize the training dynamics. Comprehensive experiments demonstrate the effectiveness of the proposed method. © 2025 SPIE.

关键词： Image fusion Image quality Feature fusion Education and training Image restoration Convolution data modeling Image enhancement Gallium nitride Semantics

来源：评论

学校读者我要写书评

暂无评论

AFRLNet: Adaptive-context Feature Representation Learning for Facial Expressions recognition via Transformers 4

AFRLNet: Adaptive-context Feature Representation Learning fo...

引用

4th International Conference on Electronic Information Engineering and computer Communication, EIECC 2024

作者： Liu, Xionghua He, Lulu Liu, Tingting Liu, Hai He, Dingxin Huazhong University of Science and Technology School of Artificial Intelligence and Automation Wuhan China Wuhan Technology and Business University School of Computer Science and Automation Wuhan China

ISBN: (纸本)9798331534622

Facial expression recognition (FER) is a critical task in intelligent educational technologies, enabling adaptive learning environments to respond to learners' emotional states and support more personalized instruction. This article introduces AFRLNet, a novel FER framework that integrates convolutional neural networks (CNNs) and transformer architectures to enhance the identification and interpretation of student facial expressions. The proposed framework includes two key components: ContextAware Local Feature Learning (CALL) and Adaptive Feature Relationship Induction (AFRI). CALL leverages attention mechanisms to emphasize symmetrical and contextually meaningful facial regions relevant to students' affective responses, while AFRI employs a Transformer-based approach to capture global features and model long-range dependencies, thus addressing challenges such as facial occlusions or varied head poses commonly found in classroom or remote learning settings. Extensive evaluations on the RAF-DB and KDEF datasets demonstrate that AFRLNet achieves state-of-the-art accuracy of 89.44% and 94.76%, respectively, surpassing existing methods. © 2024 IEEE.

关键词： Convolutional neural networks

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：