检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Chen, Zhuo Shi, Fei Zhao, Qi Institute for Interdisciplinary Information Sciences Tsinghua University Beijing100084 China QICI Quantum Information and Computation Initiative School of Computing and Data Science The University of Hong Kong Pokfulam Road Hong Kong

quantum entanglement and nonlocality are foundational to quantum technologies, driving quantum computation, communication, and cryptography innovations. To benchmark the capabilities of these quantum techniques, efficient detection and accurate quantification methods are indispensable. This paper focuses on the concept of "detection length"–—a metric that quantifies the extent of measurement globality required to verify entanglement or nonlocality. We extend the detection length framework to encompass various entanglement categories and nonlocality phenomena, providing a comprehensive analytical model to determine detection lengths for specified forms of entanglement. Furthermore, we exploit semidefinite programming techniques to construct entanglement witnesses and Bell’s inequalities tailored to specific minimal detection lengths, offering an upper bound for detection lengths in given states. By assessing the noise robustness of these witnesses,we demonstrate that witnesses with shorter detection lengths can exhibit superior performance under certain conditions. © 2024, CC BY.

关键词： quantum entanglement

来源：评论

学校读者我要写书评

暂无评论

AI-based Framework for Discriminating Human-authored and AI-generated Text

IEEE Transactions on Artificial Intelligence

引用

IEEE Transactions on Artificial Intelligence 2024年

作者： Wani, Mudasir Ahmad Abd El-Latif, Ahmed A. Elaffendi, Mohammad Hussain, Amir Prince Sultan University EIAS Data Science Lab College of Computer and Information Sciences Center of Excellence in Quantum and Intelligent Computing Riyadh11586 Saudi Arabia Menoufia University Department of Mathematics and Computer Science Faculty of Science Shebin El-Koom32511 Egypt Edinburgh Napier University School of Computing Merchiston Campus EdinburghEH10 5DT United Kingdom

Deep learning techniques are increasingly adept at distinguishing between human-written and AI-generated text. This study presents a deep learning-based text classification system to discern the source of text - whether it's generated by AI or authored by a human. We compiled a comprehensive and diverse dataset incorporating existing datasets, real-time tweets, and synthetic data generated by cutting-edge AI models such as Google Bard and OpenAI's ChatGPT. Seven deep learning techniques including Convolutional Neural Networks (CNN), Artificial Neural Networks (ANN), Recurrent Neural Networks (RNN), Long Short-Term Memory networks (LSTM), Gated Recurrent Unit (GRU), Bidirectional LSTM (BiLSTM), and Bidirectional RNN (BiRNN) and four text embedding techniques namely One-hot, Word2Vec, TF-IDF, and BERT, are employed to perform 28 experiments (4∗7). Our findings exhibit outstanding performance across all models. BERT-based models notably outshine the others, with BERT-CNN achieving the highest accuracy of 98.87%, an MCC of 0.9742, and an F1 measure of 0.9916. Following closely, Bert-BiLSTM attained an accuracy, MCC, and F1 score of 98.77%, 0.9720, and 0.9909, respectively. This paper demonstrates the effectiveness of deep learning approaches in distinguishing between human and AI-generated text, with potential applications in automated content moderation, detection of AI-generated spam, and safeguarding the authenticity of user-generated content. © 2020 IEEE.

关键词： Convolutional neural networks

来源：评论

学校读者我要写书评

暂无评论

Query-Efficient Video Adversarial Attack with Stylized Logo

arXiv

引用

arXiv 2024年

作者： Tang, Duoxun Cao, Yuxin Xiao, Xi Wang, Derui Wen, Sheng Zhu, Tianqing The College of Science Sichuan Agricultural University Ya’an China The School of Computing National University of Singapore Singapore The Tsinghua Shenzhen International Graduate School Tsinghua University Shenzhen China The Cybersecurity and Quantum Systems Group CSIRO’s Data61 Australia The School of Science Computing and Engineering Technologies Swinburne University of Technology MelbourneVIC Australia The Faculty of Data Science City University of Macau China

Video classification systems based on Deep Neural Networks (DNNs) have demonstrated excellent performance in accurately verifying video content. However, recent studies have shown that DNNs are highly vulnerable to adversarial examples. Therefore, a deep understanding of adversarial attacks can better respond to emergency situations. In order to improve attack performance, many style-transfer-based attacks and patch-based attacks have been proposed. However, the global perturbation of the former will bring unnatural global color, while the latter is difficult to achieve success in targeted attacks due to the limited perturbation space. Moreover, compared to a plethora of methods targeting image classifiers, video adversarial attacks are still not that popular. Therefore, to generate adversarial examples with a low budget and to provide them with a higher verisimilitude, we propose a novel black-box video attack framework, called Stylized Logo Attack (SLA). SLA is conducted through three steps. The first step involves building a style references set for logos, which can not only make the generated examples more natural, but also carry more target class features in the targeted attacks. Then, reinforcement learning (RL) is employed to determine the style reference and position parameters of the logo within the video, which ensures that the stylized logo is placed in the video with optimal attributes. Finally, perturbation optimization is designed to optimize perturbations to improve the fooling rate in a step-by-step manner. Sufficient experimental results indicate that, SLA can achieve better performance than state-of-the-art methods and still maintain good deception effects when facing various defense methods. We believe that the proposed SLA can raise awareness among the security community about the reliability and security of video classification systems and serve as a memorandum of possible attack methods for the future. Copyright © 2024, The Authors. All rights

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

Boosting Meaningful Dependency Mining with Clustering and Covariance Analysis

Boosting Meaningful Dependency Mining with Clustering and Co...

引用

International Conference on data Engineering

作者： Xi Wang Ruochun Jin Wanrong Huang Yuhua Tang Institute for Quantum Information & State Key Laboratory of High Performance Computing College of Computer Science and Technology National University of Defense Technology Changsha China Department of Intelligent Data Science College of Computer Science and Technology National University of Defense Technology Changsha China

ISBN: (数字)9798350317152

ISBN: (纸本)9798350317169

Functional dependencies (FDs) form a valuable ingredient for various data management tasks. However, existing methods can hardly discover practical and interpretable FDs, especially in large noisy real-life datasets. This paper studies the problem of discovering meaningful functional dependencies (FDms) that utilize support and error parameters to capture interesting dependencies in such datasets and proposes an efficient discovery algorithm called FDMε. In order to scale with large datasets, FDM ε employs an efficient sampling method with accuracy guarantees to capture the differences between tuple pairs and to quantify the connection between support/error of dependencies on samples and those on the entire dataset. Moreover, it adopts a clustering-based correlated attributes extraction to divide the exponentially large search space into multiple small sub-spaces and proposes an easy-first traversal strategy with covariance-based guidance that quickly detects candidate dependencies and validates them. Additionally, we prove a covariance lower bound as an additional pruning criterion to reduce the search space. Extensive experiments on real-life and synthetic datasets demonstrate that FDM ε is 14 times faster than existing discovery algorithms on average, up to 31 times, and scales to larger datasets with the least memory cost.

关键词： Accuracy Costs Clustering algorithms Search problems data engineering Noise measurement Frequency division multiplexing

来源：评论

学校读者我要写书评

暂无评论

Error Interference in quantum Simulation

arXiv

引用

arXiv 2024年

作者： Chen, Boyang Xu, Jue Zhao, Qi Yuan, Xiao Department of Computer Science and Technology Tsinghua University Beijing100084 China QICI Quantum Information and Computation Initiative School of Computing and Data Science The University of Hong Kong Pokfulam Road Hong Kong Center on Frontiers of Computing Studies Peking University Beijing100871 China School of Computer Science Peking University Beijing100871 China

Understanding algorithmic error accumulation in quantum simulation is crucial due to its fundamental significance and practical applications in simulating quantum many-body system dynamics. Conventional theories typically apply the triangle inequality to provide an upper bound for the error. However, these often yield overly conservative and inaccurate estimates as they neglect error interference — a phenomenon where errors in different segments can destructively interfere. Here, we introduce a novel method that directly estimates the long-time algorithmic errors with multiple segments, thereby establishing a comprehensive framework for characterizing algorithmic error interference. We identify the sufficient and necessary condition for strict error interference and introduce the concept of approximate error interference, which is more broadly applicable to scenarios such as power-law interaction models, the Fermi-Hubbard model, and higher-order Trotter formulas. Our work demonstrates significant improvements over prior ones and opens new avenues for error analysis in quantum simulation, offering potential advancements in both theoretical algorithm design and experimental implementation of Hamiltonian simulation. Copyright © 2024, The Authors. All rights reserved.

关键词： Hamiltonians

来源：评论

学校读者我要写书评

暂无评论

Sparse Channel Estimation in Massive MIMO Systems Using Compressed Sensing and Neural Networks

Sparse Channel Estimation in Massive MIMO Systems Using Comp...

引用

Intelligent Signal Processing and Effective Communication Technologies (INSPECT), IEEE International Conference on

作者： Deepak Upadhyay Kunj Bihari Sharma Mridul Gupta Abhay Upadhyay Nookala Venu Dept. of Computer Science and Engineering Graphic Era Hill University Dehradun India Dept. of Computer Science and Engineering Graphic Era deemed to be University Dehradun India Data Science and Quantum Computing ABV Indian Institute of Information Technology and Management Gwalior India

ISBN: (数字)9798350379525

ISBN: (纸本)9798350379532

In this paper, based on compressed sensing and neural networks, we present an intelligent sparse channel estimation technique for Massive MIMO systems. We show that our approach improves considerably the estimation accuracy in terms of Mean Squared Error (MSE) and Normalized Mean Squared Error (NMSE) with respect to classical methods. We demonstrate [Figure: ] that our approach is computationally more efficient, requires less time and resources (memory/disk) compared to the existing literature but for a robustness of performance across distal or low Signal-to-Noise Ratios (SNRs). We also confirm our method exhibits fast convergence and efficiently accommodates a variety of sparsity levels. By visualizing it using heatmaps and ROC curves we show that this method is able to effectively differentiate between signal and noise, asserting our approach as a powerful tool for the contemporary set of problems present in wireless communications.

关键词： Wireless communication Heating systems Accuracy Neural networks Channel estimation Estimation Massive MIMO Robustness Compressed sensing Convergence

来源：评论

学校读者我要写书评

暂无评论

Transfer Learning Approaches for Channel State Prediction in Heterogeneous Fading Networks

Transfer Learning Approaches for Channel State Prediction in...

引用

Intelligent Signal Processing and Effective Communication Technologies (INSPECT), IEEE International Conference on

ISBN: (数字)9798350379525

ISBN: (纸本)9798350379532

In this work, we leverage state-of-the-art machine learning algorithms to predict Channel State Information (CSI), for enhancing performance of wireless communication systems. The simulation and analysis stop with transfer learning to finetune, adaptive domain adaptations, multitasking (MTL) as well a few shot meta-learning models framework under different network packet loss rate scenarios. By generalizing better from one domain to another, transfer learning pushes the performance beyond what traditional methods can obtain. Further, fine-tuning also helps to align the source and target domain distributions in feature space by a partial (or full) reduction of re-stylization error. This work proposes multi-Task learning as a tool to mitigate loses in light of predicting across fading scenarios. Considering multi-task learning also provides benefits from shared learnings, so experience is pooled between the tasks which indirectly improve prediction as well and it helps in training both of these algorithms together. Moreover, the few-shot learning models prove to be highly desirable as they deliver state-of-the-art performance with minimal training data for difficult problems. These results show the potential of AI-based methods to improve CSI prediction, and have implications for improving reliability and throughput when deploying real wireless communication systems.

关键词： Wireless communication Fading channels Machine learning algorithms Transfer learning Training data Predictive models Multitasking data models Few shot learning Channel state information

来源：评论

学校读者我要写书评

暂无评论

A Detailed Analysis of Secrecy and Performance Metrics in Hybrid RF/ FSO Communication Systems: Implications of Composite Weibull and Log-normal Fading with M Turbulent Channels

A Detailed Analysis of Secrecy and Performance Metrics in Hy...

引用

Intelligent Signal Processing and Effective Communication Technologies (INSPECT), IEEE International Conference on

ISBN: (数字)9798350379525

ISBN: (纸本)9798350379532

In this paper, we have analyzed the secrecy performance and system operation performance for hybrid RF/FSO systems against composite Weibull-log-normal fading channel with M turbulent channels. The performance is evaluated by carrying out extensive simulations to calculate Signal-to-Noise Ratio (SNR), Bit Error Rate (BER) and outage probability, we also re-evaluate channel capacity and secrecy capacity with respect of environmental conditions. Compared with RF only system, FSO only standalone scheme and some meaningful orthogonal-modulation multicarrier IM/DD-based schemes proposed by the literatures. Similarly, the utilization of adaptive modulation schemes improved resistance to fading and roughness in systems. The above results confirm the hybridization of both RF and FSO a comprehensive solution for dependable communication in different and even hostile surrounding conditions which set this work as reliable foundation point towards forthcoming research activity or application relevance within all range telecommunication sector.

关键词： Fading channels Radio frequency Communication systems Systems operation Simulation Bit error rate Modulation Probability Power system reliability Signal to noise ratio

来源：评论

学校读者我要写书评

暂无评论

SimSIMS: Simulation-based Supernova Ia Model Selection with thousands of latent variables

arXiv

引用

arXiv 2023年

作者： Karchev, Konstantin Trotta, Roberto Weniger, Christoph Theoretical and Scientific Data Science SISSA Trieste Italy Department of Physics Imperial College London United Kingdom Italian Research Center on High Performance Computing Big Data and Quantum Computing Italy National Institute for Nuclear Physics Trieste Italy GRAPPA Institute University of Amsterdam Netherlands

We present principled Bayesian model comparison through simulation-based neural classification applied to SN Ia analysis. We validate our approach on realistically simulated SN Ia light curve data, demonstrating its ability to recover posterior model probabilities while marginalizing over > 4000 latent variables. The amortized nature of our technique allows us to explore the dependence of Bayes factors on the true parameters of simulated data, demonstrating Occam’s razor for nested models. When applied to a sample of 86 low-redshift SNæ Ia from the Carnegie Supernova Project, our method prefers a model with a single dust law and no magnitude step with host mass, disfavouring different dust laws for low- and high-mass hosts with odds in excess of 100:1. Copyright © 2023, The Authors. All rights reserved.

关键词： Supernovae

来源：评论

学校读者我要写书评

暂无评论

Stock Market Prediction Using Machine Learning Based Techniques

Stock Market Prediction Using Machine Learning Based Techniq...

引用

Intelligent Signal Processing and Effective Communication Technologies (INSPECT), IEEE International Conference on

作者： Aditya Dubey Ramnaresh Sharma Dhananjay Bisen Nookala Venu Varshali Jaiswal Bhagat Singh Raghuwanshi Khemchand Shakywar Centre for Internet of Things Madhav Institute of Technology & Science Gwalior India Centre for AI Madhav Institute of Technology & Science Gwalior India Data Science and Quantum Computing ABV Indian Institute of Information Technology and Management Gwalior India School of Engineering Avantika University Ujjain MP

ISBN: (数字)9798350379525

ISBN: (纸本)9798350379532

Many methods have been used to forecast stock market trends in the big data age, including real numbers, fuzzy time series data, fuzzy sets design, and conventional time series data. The control of semantic value data is made possible by the use of fuzzy time series data in stock market prediction, which makes accurate value predictions possible. These techniques are mostly used to forecast dynamic, non-linear datasets in a variety of industries, such as stock markets and cryptocurrencies. In order to predict stock prices, artificial intelligence (AI) methods such as neural networks (NN) have become increasingly important. Many methods have been developed to validate parametric and non-parametric theories for predicting trends and gains in the stock market using the feed-forward (FF) NN technique. A variety of soft computing techniques have been investigated in order to provide ways for decision-making based on profit and loss criteria without relying on artificial intelligence. This paper's main goal is to perform a comprehensive examination of the different prediction techniques used to estimate price and return in the stock market. The study includes a comparative review of several approaches as well as a variety of stock market forecasting strategies.

关键词： Computational modeling Time series analysis Signal processing algorithms Predictive models Prediction algorithms Market research Vectors Stock markets Forecasting Investment

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：