检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Ma, Shixuan Wang, Quan MOE Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and Telecommunications China

The increasing capability and widespread usage of large language models (LLMs) highlight the desirability of automatic detection of LLM-generated text. Zero-shot detectors, due to their training-free nature, have received considerable attention and notable success. In this paper, we identify a new feature, token cohesiveness, that is useful for zero-shot detection, and we demonstrate that LLM-generated text tends to exhibit higher token cohesiveness than human-written text. Based on this observation, we devise TOCSIN, a generic dual-channel detection paradigm that uses token cohesiveness as a plug-and-play module to improve existing zero-shot detectors. To calculate token cohesiveness, TOCSIN only requires a few rounds of random token deletion and semantic difference measurement, making it particularly suitable for a practical black-box setting where the source model used for generation is not accessible. Extensive experiments with four state-of-the-art base detectors on various datasets, source models, and evaluation settings demonstrate the effectiveness and generality of the proposed approach. Code available at: https://***/Shixuan-Ma/TOCSIN. © 2024, CC BY-NC-ND.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Improving Knowledge Distillation for Federated Learning on Non-IID Data

Improving Knowledge Distillation for Federated Learning on N...

引用

IEEE International Conference on Big Data

作者： Zongyi Chen Sanchuan Guo Liyan Shen Xi Zhang Zhuonan Chang Key Laboratory of Trustworthy Distributed Computing and Service (MOE) Beijing University of Posts and Telecommunications China

Federated learning (FL) leverages knowledge from decentralized clients to train a global model in a privacy-preserving manner. One of the critical challenges in FL is the heterogeneity of client local data (i.e. non-IID data), which can lead to significant performance degradation. Federated Distillation (FD) can mitigate this issue by distilling the client predictions on the unlabeled public data into a global student model. However, a simple averaging of the client predictions without considering model confidences may lead to erroneous predictions. Moreover, it is challenging to rectify the predicted pseudo labels as no human labels are available in the public data. To address these issues, we propose a novel FD framework named FedUSL, aiming to distill a powerful global model on non-IID data. Specifically, we estimate the uncertainty of each client’s prediction on the unlabeled data and adopt a weighted ensemble of their predictions in consideration of uncertainties. We further rectify the global model predictions by a self-label reassigning method, without the requirement of manual labels. Extensive experiments on image and text tasks show that our proposal can achieve superior performance than state-of-the-art methods, and incurs no extra computing burden on the client side. The code is available at https://***/r/FedUSL-6338/.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Real Time Attrition Prediction Mechanism Based on Deep Learning

Real Time Attrition Prediction Mechanism Based on Deep Learn...

引用

Signal Processing, Communications and Engineering Management Conference (ISPCEM), International

作者： Chen Wu Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and Telecommunications Beijing China

Modern companies mostly have their own employee management systems, which retain a large amount of employee behavior data. These data can be used to mine valuable information, such as whether employees have attrition intentions. This paper proposes a mechanism of attrition prediction based on deep learning, which mainly includes three modules: (1) Using LSTM to build a performance-attrition pre-trained model; (2)Using Transformer to build a sentiment tendency model based on employee weekly reports; (3) Based on (1), (2), establishing Real Time Attrition Prediction Mechanism in the company's employee management system to predict employees' willingness to leave at any time. The purpose of this mechanism is to help companies manage better and reduce losses caused by employee attrition.

关键词： Deep learning Companies Predictive models Signal processing Transformers Real-time systems Behavioral sciences

来源：评论

学校读者我要写书评

暂无评论

A Lightweight Bit-Operation Abnormal Traffic Detection Method Based On XNOR-CNN

A Lightweight Bit-Operation Abnormal Traffic Detection Metho...

引用

IEEE Conference on Wireless Communications and Networking

作者： Yueqin Ge Xiaoyong Li Binsi Cai Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and Telecommunications Beijing China

The rapid development of the Internet and the increasingly complex structure of network space make the network security situation more and more serious. Abnormal traffic detection is an important part of network intrusion detection and plays an important role in the field of network security. Traditional machine learning methods rely too much on feature selection and extraction and have high false positive rate and delay caused by abnormal behavior recognition. Deep learning models such as CNN and RNN have many parameters and long training time, which seriously affect the early warning function of network intrusion detection. Aiming at the problem that the parameter redundancy of deep learning models limits the corresponding model deployment in some scenarios and devices, this paper proposes a lightweight bit-operation abnormal traffic detection method based on XNOR-net convolutional neural network (XNOR-CNN). The model uses convolutional neural network(CNN) and long-short term memory(LSTM) to comprehensively analyze the temporal and spatial characteristics of network traffic, and predicts and classifies the attack behaviors in the future network traffic. In particular, the model transforms the complex convolution process into bit operation between vectors by XNOR operation, which reduces the storage of weight vectors and complex redundant calculation, greatly improves the speed of network training and reduces the consumption of network memory. In this paper, the accuracy and efficiency of XNORCNN model are proved through a large number of comparative experiments on real traffic datasets.

关键词： Deep learning Training Network intrusion detection Telecommunication traffic Predictive models Network security Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Towards Accurate and Efficient Super Spreader Detection with Sketching 8

Towards Accurate and Efficient Super Spreader Detection with...

引用

8th International Conference on Data Science in Cyberspace, DSC 2023

作者： Song, Shouyou Wang, Pu Li, Yangchun Wang, Nanyang Jiang, Wei Bupt Key Laboratory of Trustworthy Distributed Computing and Service Ministry of Education Beijing China Chinese Academy of Cyberspace Studies Beijing China Bupt Beijing China School of Computer and Information Engineering Chuzhou University Anhui China

ISBN: (纸本)9798350331035

Super spreaders are the flow that have a large number of distinct connections (also called spread), which related with many threats to networks. Estimating flow spread is the crucial step in super spreader detection. However, existing methods cannot achieve flow spread estimation in terms of accurate, efficient, and reversible simultaneously. All these characteristics is highly required for high-speed network measurement. In this paper, we propose MorphSketch, a new data structure that estimates flow spread for super spreader detection with high accuracy, memory efficiency, high throughput and reversibility. MorphSketch combines hashing with sampling to process packets in order to improve throughput. It uses self-morph bitmap to record spread information, which can adaptively enlarge the upper bound of spread estimation under limited memory usage to ensure accuracy and memory efficiency. Moreover, MorphSketch can track candidate super spreader by comparing corresponding spread information, which realizes reversibility in super spreader detection. We perform a series of performance evaluations on real world traffic trace. Experiment results demonstrate that under same memory usage, the MorphSketch significantly outperforms existing work in terms of accuracy and efficiency. © 2023 IEEE.

关键词： bitmap sketch Spread estimation super spreader detection

来源：评论

学校读者我要写书评

暂无评论

QPSK pilot-aided carrier phase recovery algorithm for 800-Gb/s/? DP-256QAM transmission

QPSK pilot-aided carrier phase recovery algorithm for 800-Gb...

引用

2020 Asia Communications and Photonics Conference, ACPC 2020/International Conference on Information Photonics and Optical Communications, IPOC 2020

作者： Cui, Han Xu, Xuekai Liu, Shuangyue Lu, Yueming Qiao, Yaojun State Key Laboratory of Information Photonics and Optical Communications School of Information and Communication Engineering Beijing University of Posts and Telecommunications Beijing100876 China Key Laboratory of Trustworthy Distributed Computing and Service Ministry of Education School of Cyberspace Security Beijing University of Posts and Telecommunications Beijing China

A QPSK pilot-aided carrier phase recovery algorithm is investigated for 800-Gb/s/? DP-256QAM transmission systems. Simulation results show that, compared with BPS, this proposed algorithm reduces by 99% computational ... 详细信息

ISBN: (纸本)9781557528209

关键词： Quadrature phase shift keying

来源：评论

学校读者我要写书评

暂无评论

APPLIED RESEARCH ON VISUAL MINING TECHNOLOGY IN MEDICAL DATA

APPLIED RESEARCH ON VISUAL MINING TECHNOLOGY IN MEDICAL DATA

引用

2016 4th IEEE International Conference on Cloud computing and Intelligence Systems（IEEE CCIS2016）

作者： Chunshan Wu Yongqiang Weng Qiaowei Jiang Wenming Guo Cong Wang Key Laboratory of Trustworthy Distributed Computing and Service Ministry of Education School of Software Engineering Beijing University of Post and Telecommunications Key Laboratory of Trustworthy Distributed Computing and Service Ministry of Education School of AutomationBeijing University of Post and Telecommunications

ISBN: (纸本)9781509012572

Data mining is a process of discovering information from large datasets by using automated algorithms, which makes it difficult for users to intuitively understand, explore and optimize the datasets and algorithms process. Recently with the development of the visual field, researchers have been exploring how to use the visualization methods to make users intuitively understand datasets and algorithms. In this paper we introduce the concept of visual data mining and dissect the visual data mining technology, mainly focusing on three related aspects: source data, data mining processes and data mining results;then design and implement a visual mining system based on decision tree algorithm, which achieves the encapsulation of data mining algorithms, and presents to the user a simple interactive interface, so that users can carry out visual data mining;finally apply the visual data mining technology to the survey data of elderly anxiety patients, and analyze the decision rules of anxiety disorders, which has important significance for clinical research and decision-making.

关键词： visualization data mining medical data analysis

来源：评论

学校读者我要写书评

暂无评论

Research on discrete emotion classification of Chinese online product reviews based on OCC model 3

Research on discrete emotion classification of Chinese onlin...

引用

3rd IEEE International Conference on Data Science in Cyberspace, DSC 2018

作者： Liu, Lina Qi, Jiayin School of Economics and Management Key Laboratory of Trustworthy Distributed Computing and Service School of Ethnic Minority Education Beijing University of Posts and Telecommunication Beijing China School of Management Shanghai University of International Business and Economics Shanghai China

ISBN: (纸本)9781538642108

There are many rich and multi-dimensional discrete emotional information in online reviews. The research of discrete emotion analysis in English online reviews is relatively early and has made some achievements, but the related research on Chinese online reviews, especially in consumption field is still in its infancy, which is the focus of this study. By taking the product reviews of a large Chinese e-commerce platform as the research object, and using the OCC model which is a kind of emotional cognition model as the basis of emotion classification, the study constructs six datasets annotated with six discrete emotions including satisfaction, disappointment, admiration, reproach, like and hate. CHI-based feature selection is performed for each review sample of the above six datasets and five traditional classifiers are used to classify the discrete emotions. The experimental result demonstrates the effectiveness of our proposed approach and shows that the classifier based on LR algorithm has the best classification effect for each discrete emotion. © 2018 IEEE.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Knowledge-based information security risk assessment method

引用

Journal of China Universities of Posts and Telecommunications 2013年第SUPPL -2期20卷 60-63页

作者： Guan, Ji-Zheng Lei, Ming-Tao Zhu, Xiao-Lu Liu, Jian-Yi GDS Services Ltd. Beijing 100016 China Automation School Beijing University of Posts and Telecommunications Beijing 100876 China School of Computer Science Beijing University of Posts and Telecommunications Beijing 100876 China Key Laboratory of Trustworthy Distributed Computing and Service Beijing University of Posts and Telecommunications Beijing 100876 China

It is an important function for managers to keep away from information security risks. With the increasing complex and scale of information systems, information system security risks may be more difficult to assess and strategies for risk reduction may be lack of objectivity. To solve this problem, this paper proposes a knowledge-based information security risk assessment method in which basic rules and specific rules are defined to match every asset, threat and vulnerability. Basic rules are defined as the rules without influence of external relationships. Specific rules are defined as the rules by user group. Performance analysis shows this method could increase efficiency and ensure accuracy of risk assessment. © 2013 The Journal of China Universities of Posts and Telecommunications.

关键词： Risk assessment

来源：评论

学校读者我要写书评

暂无评论

RBCP: A Reputation-based Blockchain Consensus Protocol for Internet of Things 24

RBCP: A Reputation-based Blockchain Consensus Protocol for I...

引用

24th IEEE International Conference on High Performance computing and Communications, 8th IEEE International Conference on Data Science and Systems, 20th IEEE International Conference on Smart City and 8th IEEE International Conference on Dependability in Sensor, Cloud and Big Data Systems and Application, HPCC/DSS/SmartCity/DependSys 2022

作者： Shang, Jiaze Lu, Tianbo Cai, Yingjie Luo, Shuang Jin, Zhaoxin School of Computer Science National Pilot Software Enginerring School Beijing University of Posts and Telecommunications Beijing China Beijing University of Posts and Telecommunications Key Laboratory of Trustworthy Distributed Computing and Service Ministry of Education Beijing China

ISBN: (纸本)9798350319934

The application of blockchain in the Internet of Things (IoT) solves the problem of centralization, enables data interaction of IoT nodes in untrusted environments. However, most of the blockchain consensus protocols are computationally expensive and poorly scalable, which cannot meet the current demand for low power consumption and low latency of IoT devices. Meanwhile, IoT nodes suffer from many malicious attacks. An effective reputation evaluation scheme is valuable for establishing a trusted, secure IoT environment with certain applications. In this paper, we propose RBCP, a novel reputation-based blockchain consensus protocol, and apply it to the IoT. The protocol divides IoT nodes into multiple regions, and consensus nodes in each region are selected based on real-time updated reputation values, which are used to participate in light PoW consensus and finally propose blocks. Then, we model the deposits and profits of the nodes as Stackelberg game and analyze the relationship between them. To evaluate our proposed protocol, we compare it with existing consensus protocols through several experiments. Our experimental results show that RBCP has advantages in terms of TPS, consensus time, and communication consumption, proving that it has low latency and less consensus time under the premise of sufficient security. © 2022 IEEE.

关键词： Internet of things

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：