检索结果-内蒙古大学图书馆

Improvement of Min-Entropy Evaluation Based on Pruning and Quantized Deep Neural Network

IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 2023年 18卷 1410-1420页

作者： Li, Haohao Zhang, Jianguo Li, Zhihu Liu, Juan Wang, Yuncai Taiyuan Univ Technol Coll Phys & Optoelect Key Lab Adv Transducers & Intelligent Control Syst Minist Educ China Taiyuan 030024 Peoples R China China Elect Power Res Inst Beijing 100192 Peoples R China Nations Technol Co Ltd Shenzhen 518057 Peoples R China Guangdong Univ Technol Adv Inst Photon Technol Guangzhou 510006 Peoples R China

In the field of information security, the unpredictability of random numbers plays determinant role according to the security of cryptographic systems. However, limited by the capability of pattern recognition and data mining, statistical-based methods for random number security assessment can only detect whether there are obvious statistical flaws in random sequences. In recent years, some machine learning-based techniques such as deep neural networks and prediction-based methods applied to random number security have exhibited superior performance. Concurrently, the proposed deep learning models bring out issues of large number of parameters, high storage space occupation and complex computation. In this paper, for the challenge of random number security analysis: building high-performance predictive models, we propose an effective analysis method based on pruning and quantized deep neural network. Firstly, we train a temporal pattern attention-based long short-term memory (TPA-LSTM) model with complex structure and good prediction performance. Secondly, through pruning and quantization operations, the complexity and storage space occupation of the TPA-LSTM model were reduced. Finally, we retrain the network to find the best model and evaluate the effectiveness of this method using various simulated data sets with known min-entropy values. By comparing with related work, the TPA-LSTM model provides more accurate estimates: the relative error is less than 0.43%. In addition, the model weight parameters are reduced by more than 98% and quantized to 2 bits (compression over 175x) without accuracy loss.

关键词： Security Predictive models Entropy Deep learning data models Quantization (signal) Neural networks Min-entropy TPA-LSTM random number pruning quantization

来源：评论

学校读者我要写书评

暂无评论

Evaluation of machine learning Models for Predicting Exports of India 7

Evaluation of Machine Learning Models for Predicting Exports...

引用

7th international conference on Computing in Engineering and Technology, ICCET 2022

作者： Gupta, Vaishali Kumar, Ela Department of Computer Science & Engineering Indira Gandhi Delhi Technical University for Women Delhi India

ISBN: (纸本)9781839537042

Analysing economic trends and predicting future value of economic variables are key factors for assessing a country's economic outlook. This study pursues a machine learning approach to predict country's exports using other macroeconomic variables, namely, country's GDP, world's exports and real effective exchange rate (REER) by deploying various machine learning algorithms. In this study, Linear Regression, Support Vector machine, Decision Tree and Random Forest machine learning models are applied on India's trade data for the period 1970-2019. By comparing the prediction accuracy, it is found that decision tree regression outperformed the other three algorithms and produced better results for predicting India's exports. © 2022 The Institution of Engineering and Technology.

关键词： Forecasting

来源：评论

学校读者我要写书评

暂无评论

Digitization of Written Parliamentary Questions from the Historical Archive (1974-1977) of the Hellenic Parliament

Digitization of Written Parliamentary Questions from the His...

引用

18th international conference on Document Analysis and Recognition (ICDAR)

作者： Fitsilis, Fotios Gatos, Basilis Palaiologos, Konstantinos Kaddas, Panagiotis Kyrkos, Charalambis Georgoulea, Maria-Eleni Armenakis, Yiannis Tasouli, Christina Mikros, George Rozenberg, Olivier Kiousi, Eleni Hellen Parliament Vas Sofias Ave 2-4 Athens 10021 Greece Natl Ctr Sci Res Demokritos Patriarchou Grigoriou E & Neapoleos 27 Aghia Paraskevi 15341 Greece Hellen OCR Team Aretis 5B Aghia Paraskevi 15343 Greece Hamad Bin Khalifa Univ Educ City Gate 8 Ar Rayyan Qatar

ISBN: (纸本)9783031706448;9783031706455

This article outlines the digitization process and methodology applied to the archive of parliamentary questions from the 1st Parliamentary Term (1974-1977) in the Hellenic Parliament. A collaborative pilot project involving parliament, academia, and a research center facilitated the conversion of printed material to open data. The main tasks of the project include capturing digital images, a custom Optical Character Recognition (OCR) software solution employing machine learning, and rigorous validation for accuracy of a fragmented and of variable quality polytonic corpus in a variety of modern Greek language called Katharevousa. The article discusses the approach and challenges as well as the initial results of the digitization effort, emphasizing ongoing research steps. Overall, 1,674 images were digitally processed corresponding to 1,338 questions. Following algorithmic training, character recognition accuracy is over 98.5%. Successful implementation streamlines further similar digitalization operations in the vast parliamentary archives, while enabling in-depth studies on parliamentary control in the turbulent period of the immediate post-junta era in Greece. A preliminary comparative analysis with a corpus of newer parliamentary questions (2009-2019) provides insights and incentives for the further study of the characteristics and evolution of the Greek language.

关键词： Hellenic Parliament written questions machine learning OCR parliamentary control polytonic corpus

来源：评论

学校读者我要写书评

暂无评论

Direct Demonstration-Based Imitation learning and Control for Writing Task of Robot Manipulator

Direct Demonstration-Based Imitation Learning and Control fo...

引用

IEEE international conference on Consumer Electronics (ICCE)

作者： Park, Sejun Park, Ju Hyun Lee, Sangmoon Kyungpook Natl Univ Dept Elect & Elect Engn Daegu South Korea Yeungnam Univ Dept Elect Engn Kyongsan South Korea

ISBN: (纸本)9781665491303

In this paper, we propose an imitation learning method based on a direct demonstration of robot manipulators. To track the desired position and force, we designed an impedance controller. As a result of imitation learning, the robot can be acted as intended even if the initial position is different, and be able to perform a writing task well even if a different contact force is applied to the changing environment. We propose Long Short-Term Memory (LSTM)-based imitation learning method through the demonstration data. Finally, the proposed method was verified by applying the writing task with the actual industrial robot manipulator that acts as the expert's intention for the direct demonstration.

关键词： Robot Manipulator Impedance Controller Direct Demonstration Imitation learning LSTM Writing Task

来源：评论

学校读者我要写书评

暂无评论

IntelliBeeHive: An Automated Honey Bee, Pollen, and Varroa Destructor Monitoring System 23

IntelliBeeHive: An Automated Honey Bee, Pollen, and Varroa D...

引用

23rd IEEE international conference on machine learning and Applications, ICMLA 2024

作者： Narcia-Macias, Christian I. Guardado, Joselito Rodriguez, Jocell Park, Junseong Rampersad-Ammons, Joanne Enriquez, Erik Kim, Dong-Chul University of Texas Rio Grande Valley Department of Computer Science United States University of Texas Rio Grande Valley Department of Biology United States School of Earth Environmental and Marine Sciences University of Texas Rio Grande Valley United States

ISBN: (纸本)9798350374889

Utilizing computer vision and the latest technological advancements, in this study, we developed a honey bee moni-toring system that aims to enhance our understanding of Colony Collapse Disorder, honey bee behavior, population decline, and overall hive health. The system is positioned at the hive entrance providing real-time data, enabling beekeepers to closely monitor the hive's activity and health through an account-based website. Using machine learning, our monitoring system can accurately track honey bees, monitor pollen-gathering activity, and detect Varroa mites, all without causing any disruption to the honey bees. Moreover, we have ensured that the development of this monitoring system utilizes cost-effective technology, making it accessible to apiaries of various scales, including hobbyists, com-mercial beekeeping businesses, and researchers. The inference models used to detect honey bees, pollen, and mites are based on the YOLOv7-tiny architecture trained with our data. The F1-score for honey bee model recognition is 0.95 and the precision and recall value is 0.981. For our pollen and mite object detection model F1-score is 0.95 and the precision and recall value is 0.821 for pollen. The overall performance of our IntelliBeeHive system demonstrates its effectiveness in monitoring the honey bee's activity, achieving an accuracy of 96.28 percent in tracking and our pollen model achieved a F1-score of 0.831. © 2024 IEEE.

关键词： Sugar beets

来源：评论

学校读者我要写书评

暂无评论

learning Analytics in Blended Mode During Covid-19 Pandemic 7

Learning Analytics in Blended Mode During Covid-19 Pandemic

引用

2023 7th international conference On Computing, Communication, Control And Automation, ICCUBEA 2023

作者： Tejal, R.K. Kanchana, V. Amrita Vishwa Vidyapeetham Department of Computer Science Mysuru India

ISBN: (纸本)9798350304268

Personality and behavior are great factors when considering students' learning strategies. Students tend to utilize their best capacity when the environment around them is comforting and the situation go hand in hand with their personality. learning analytics is a study based on theories and until recently, was not a topic any researcher would show interest in. During 2019- 2020, a severe change was caused not only to students but the entire education system when the Covid-19 pandemic was caused. Exploring the factors which affected students' performance during and after the pandemic is a topic yet to be researched. This entire study is based on students' feedback on their performance during and after the pandemic has nearly ended © 2023 IEEE.

关键词： Anova blended learning data mining learning analytics SVM-Support vector machine

来源：评论

学校读者我要写书评

暂无评论

Federated learning for Metaverse: A Survey

Federated Learning for Metaverse: A Survey

引用

32nd World Wide Web conference (WWW)

作者： Chen, Yao Huang, Shan Gan, Wensheng Huang, Gengsen Wu, Yongdong Jinan Univ Guangzhou Peoples R China

ISBN: (纸本)9781450394161

The metaverse, which is at the stage of innovation and exploration, faces the dilemma of data collection and the problem of private data leakage in the process of development. This can seriously hinder the widespread deployment of the metaverse. Fortunately, federated learning (FL) is a solution to the above problems. FL is a distributed machine learning paradigm with privacy-preserving features designed for a large number of edge devices. Federated learning for metaverse (FL4M) will be a powerful tool. Because FL allows edge devices to participate in training tasks locally using their own data, computational power, and model-building capabilities. Applying FL to the metaverse not only protects the data privacy of participants but also reduces the need for high computing power and high memory on servers. Until now, there have been many studies about FL and the metaverse, respectively. In this paper, we review some of the early advances of FL4M, which will be a research direction with unlimited development potential. We first introduce the concepts of metaverse and FL, respectively. Besides, we discuss the convergence of key metaverse technologies and FL in detail, such as big data, communication technology, the Internet of Things, edge computing, blockchain, and extended reality. Finally, we discuss some key challenges and promising directions of FL4M in detail. In summary, we hope that our up-to-date brief survey can help people better understand FL4M and build a fair, open, and secure metaverse.

关键词： Metaverse federated learning intelligent applications survey

来源：评论

学校读者我要写书评

暂无评论

Online machine learning for Real-Time Cell Culture Process Monitoring 37th

Online Machine Learning for Real-Time Cell Culture Process ...

引用

37th Australasian Joint conference on Artificial Intelligence, AJCAI 2024

作者： Khuat, Thanh Tung Bassett, Robert Otte, Ellen Gabrys, Bogdan Complex Adaptive Systems Laboratory The Data Science Institute University of Technology Sydney UltimoNSW2007 Australia CSL Innovation MelbourneVIC3000 Australia

ISBN: (纸本)9789819603503

In the biopharmaceutical industry, the production of monoclonal antibodies (mAbs) relies heavily on the optimization and real-time monitoring of cell culture processes. This study assesses the effectiveness of online machine learning (OML) models for the real-time monitoring of nutrient and metabolite concentrations, such as glucose, lactate, and ammonium, during the cell cultivation process using Raman spectroscopy as input data. The research addresses the challenge of limited data inherent in cell culture processes, particularly when transferring learned information to new bioreactor runs with different cell lines and media compositions. Traditional pretrained machine learning models often fail to generalize their learned information in such context. Our research demonstrates that OML models can dynamically adapt to the evolving conditions of cell culture processes, providing accurate and timely updates on critical nutrient and metabolite concentrations. We conducted a series of experiments comparing the performance of OML models against pretrained models in monitoring cell culture processes. The findings indicate that OML models exhibit superior performance in capturing the dynamic changes of nutrient and metabolite concentrations during cell cultivation. In contrast, pretrained models exhibit substantial performance degradation when applied to a completely new bioreactor run, highlighting the limitations of static learning approaches in handling the complex, dynamic nature of cell cultures. This study emphasizes the potential of OML in enhancing the robustness and efficiency of real-time process monitoring in mAb production, paving the way for more adaptive and resilient biomanufacturing processes. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2025.

关键词： Metabolites

来源：评论

学校读者我要写书评

暂无评论

Classification and detection of EEG data for tired driving based on Mamba framework 4

Classification and detection of EEG data for tired driving b...

引用

4th IEEE international conference on Digital Twins and Parallel Intelligence, DTPI 2024

作者： Xiaohui, Chen Jian, Liu University of Science and Technology Beijing Beijing China

ISBN: (纸本)9798350349252

Fatigue driving is an important factor that causes frequent traffic accidents. The fatigue driving pre-detection system is a research direction at present. Compared to the driving habits of different people, EEG signals can show physiological consistency, and it is a major challenge to mine the characteristics of driver fatigue from the multi-channel EEG data. In this paper, based on Mamba framework, CNN feature extraction module is selected to realize coupling, and a CNN-Mamba model is constructed and applied to EEG fatigue driving detection task. In this paper, we selected a fatigue driving dataset published in 2019, which contains 30 channels of EEG data from 11 subjects. The results show that the average accuracy of the model is 72.73% in the cross-subject EEG classification of 11 subjects, which is significantly improved compared with other advanced methods, and the recognition accuracy performance is improved by 9.65% compared with EEGNet. © 2024 IEEE.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Demonstration Shaped Reward machine for Robot Assembly Reinforcement learning Tasks 16th

Demonstration Shaped Reward Machine for Robot Assembly Rein...

引用

16th international conference on Intelligent Robotics and Applications, ICIRA 2023

作者： Xiao, Ruihong Zhan, Hong Jiang, Yiming Yang, Chenguang School of Automation Science and Engineering South China University of Technology Guangzhou China School of Robotics Hunan University Changsha China

ISBN: (纸本)9789819964949

Reinforcement learning has been proven to be successfully applied in robotic manipulation tasks, but it has been confronted with problems about interacting with the dynamic environment and generating high-level policies. Existing reward functions include sparse and shaped reward, and other way to view the reward functions as black boxes and then discover the best strategies by interacting with the environment. In this paper, a reward machine with demonstration definition for robot assembly is presented, which is a finite state machine that supports generating the specified reward function from the demonstration data and displays the reward function structure. By dividing the task demonstration data, extracting the key points of the task and defining the state of the robot, a reward function based on state migration is generated. The robot chooses the best reward according to the current state, which is conducive to using the reward structure to improve the sample efficiency and the quality of the result strategy. In this paper, we focus on the robot assembly reinforcement learning task. The structure of the proposed method is used to carry out experiments in the robot simulator. We compare the design methods with different reward functions and it proves that the proposed method can effectively improve the training efficiency of robot reinforcement learning. © The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd. 2023.

关键词： Reinforcement learning

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：