检索结果-内蒙古大学图书馆

International Conference on Acoustics, Speech, and Signal Processing (ICASSP)

作者： Hang Shao Bei Liu Yanmin Qian Auditory Cognition and Computational Acoustics Lab MoE Key Lab of Artificial Intelligence AI Institute Department of Computer Science and Engineering Shanghai Jiao Tong University Shanghai China

Various Large Language Models (LLMs) from the Generative Pretrained Transformer (GPT) family have achieved outstanding performances in a wide range of text generation tasks. However, the enormous model sizes have hindered their practical use in real-world applications due to high inference latency. Therefore, improving the efficiencies of LLMs through quantization, pruning, and other means has been a key issue in LLM studies. In this work, we propose a method based on Hessian sensitivity-aware mixed sparsity pruning to prune LLMs to at least 50% sparsity without the need of any retraining. It allocates sparsity adaptively based on sensitivity, allowing us to reduce pruning-induced error while maintaining the overall sparsity level. The advantages of the proposed method exhibit even more when the sparsity is extremely high. Furthermore, our method is compatible with quantization, enabling further compression of LLMs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

A neuro-observer-based optimal control for nonaffine nonlinear systems with control input saturations

引用

Control Theory and Technology 2021年第2期19卷 283-294页

作者： Behzad Farzanegan Mohsen Zamani Amir Abolfazl Suratgar Mohammad Bagher Menhaj Distributed Intelligent Optimization Research Lab Department of Electrical EngineeringAmirkabir University of TechnologyTehranIran Computational Intelligence Lab Department of Electrical EngineeringAmirkabir University of TechnologyTehranIran The School of Electrical Engineering and Computer Science The University of NewcastleNewcastleAustralia Department of Medical Physics and Engineering Shiraz University of Medical SciencesShirazIran

In this study,an adaptive neuro-observer-based optimal control(ANOPC)policy is introduced for unknown nonaffine nonlinear systems with control input ***–Jacobi–Bellman(HJB)framework is employed to minimize a non-quadratic cost function corresponding to the constrained control *** consists of both analytical and algebraic *** the analytical part,first,an observer-based neural network(NN)approximates uncertain system dynamics,and then another NN structure solves the HJB *** the algebraic part,the optimal control input that does not exceed the saturation bounds is *** weights of two NNs associated with observer and controller are simultaneously updated in an online *** ultimately uniformly boundedness(UUB)of all signals of the whole closed-loop system is ensured through Lyapunov’s direct ***,two numerical examples are provided to confirm the effectiveness of the proposed control strategy.

关键词： Input constraints Optimal control Neural networks Nonaffine nonlinear systems Reinforcement learning Unknown dynamics

来源：评论

学校读者我要写书评

暂无评论

Massive Dimensions Reduction and Hybridization with Meta-heuristics in Deep Learning

Massive Dimensions Reduction and Hybridization with Meta-heu...

引用

Canadian Conference on Electrical and Computer engineering (CCECE)

作者： Rasa Khosrowshahli Shahryar Rahnamayan Beatrice Ombuki-Berman Department of Computer Science Nature-Inspired Computational Intelligence (NICI) Lab Bio-Inspired Computational Intelligence Research Group (BICIG) Brock University St. Catharines ON Canada Department of Engineering Nature-Inspired Computational Intelligence (NICI) Lab Brock University St. Catharines ON Canada Department of Computer Science Bio-Inspired Computational Intelligence Research Group (BICIG) Brock University St. Catharines ON Canada

ISBN: (数字)9798350371628

ISBN: (纸本)9798350371635

Deep learning is mainly based on utilizing gradient-based optimization for training Deep Neural Network (DNN) models. Although robust and widely used, gradient-based optimization algorithms are prone to getting stuck in local minima. In this modern deep learning era, the SOTA DNN models have millions and billions of parameters, including weights and biases, making them huge-scale optimization problems in terms of search space. Tuning a huge number of parameters is a challenging task that causes vanishing/exploding gradients and overfitting; likewise, utilized loss functions do not exactly represent our targeted performance metrics. A practical solution to exploring large and complex solution space is meta-heuristic algorithms. Since DNNs exceed thousands and millions of parameters, even robust meta-heuristic algorithms, such as Differential Evolution, struggle to efficiently explore and converge in such huge-dimensional search spaces, leading to very slow convergence and high memory demand. To tackle the mentioned curse of dimensionality, the concept of blocking was recently proposed as a technique that reduces the search space dimensions by grouping them into blocks. In this study, we aim to introduce Histogram-based Blocking Differential Evolution (HBDE), a novel approach that hybridizes gradient-based and gradient-free algorithms to optimize parameters. Experimental results demonstrated that the HBDE could reduce the parameters in the ResNet-18 model from 11M to 3K during the training/optimizing phase by meta-heuristics, namely, the proposed HBDE, which outperforms baseline gradient-based and parent gradient-free DE algorithms evaluated on CIFAR10 and CIFAR100 datasets showcasing its effectiveness with reduced computational demands for the very first time.

关键词： Deep learning Training Measurement Dimensionality reduction Metaheuristics Memory management Artificial neural networks

来源：评论

学校读者我要写书评

暂无评论

Optimizing Moving Target Defense For Cyber Anomaly Detection 1

Optimizing Moving Target Defense For Cyber Anomaly Detection

引用

1st International Conference on computational Intelligence, Communication Technology and Networking, CICTN 2023

作者： Vubangsi, M. Mubarak, Auwalu Saleh Yayah, Jameel Altrjman, Chadi Manwal, Manika Yadav, Satya Prakash Al-Turjman, Fadi Research Center for Ai and IoT Ai and Robotics Institute Near East University Ai Engineering Dep't Mersin10 Turkey Htttc Bambili University of Bamenda Computational Materials Science Lab NWR Bambili Cameroon Waterloo University Canada GraphicEra Hill University Computer Science and Engineering Dehradun India Affiliated to Aktu Greater Noida India University of Kyrenia Faculty of Engineering Mersin10 Turkey

ISBN: (纸本)9798350338027

In this research, we evaluate the effectiveness of different MTD techniques on the transformer-based cyber anomaly detection models trained on the KDD Cup'99 Dataset, a publicly available dataset commonly used for evaluating intrusion detection systems. We explore the trade-offs between security and performance when using MTD techniques for cyber anomaly detection and investigate how MTD techniques can be combined with other cybersecurity techniques to improve the overall security of the system. We evaluate their performance using standard metrics such as accuracy and FI score, as well as measures of robustness against adversarial attacks. Our results show that MTD techniques can significantly improve the security of the anomaly detection model, with some techniques being more effective than others depending on the model architecture. We also find that there are trade-offs between security and performance, with some MTD techniques leading to a reduction in model accuracy or an increase in computation time. However, we demonstrate that these tradeoffs can be mitigated by optimizing the MTD parameters for the specific model architecture. © 2023 IEEE.

关键词： Cybersecurity

来源：评论

学校读者我要写书评

暂无评论

A hybrid lithium sulfonated polyoxadiazole derived single-ion conducting gel polymer electrolyte enabled effective suppression of dendritic lithium growth

引用

Chinese Chemical Letters 2022年第2期33卷 1025-1031页

作者： Dazhe Li Longbo Luo Jiadeng Zhu Haimei Qin Pengqing Liu Zhaomei Sun Yi Lei Mengjin Jiang College of Polymer Science&Engineering Sichuan UniversityChengdu 610065China Chemical Sciences Division Oak Ridge National LaboratoryOak RidgeTN 37831United States State Key Laboratory for Physical Chemistry of Solid Surfaces Fujian Provincial Key Lab of Theoretical and Computational Chemistryand College of Chemistry and Chemical EngineeringXiamen UniversityXiamen 361005China

Lithium metal is deemed as an ideal anode material in lithium-ion batteries because of its ultrahigh theoretical specific capacity and the lowest redox ***,the rapid capacity attenuation and inferior security resulting from the dendritic lithium growth severely limit its *** a novel hybrid gel polymer electrolyte (GPE) based on electrospun lithium sulfonated polyoxadiazole (LiSPOD) nanofibrous membrane swelled by lithium bis(trifluoromethanesulfonyl)imide (Li TFSI) ether liquid electrolyte is proposed to address the issue of lithium *** Li-SPOD membrane synthesized by a simple one-pot method exhibits excellent mechanical strength and thermal resistance due to its high molecular weight and rigid *** electron-withdrawing oxadiazole ring and oxadiazole ring-Li;complex,and N,O heteroatoms with lone pairs of electrons in Li-SPOD macromolecular chains facilitate the dissociation of-SO_(3)Li group and Li^(+)*** hybrid Li-SPOD GPE exhibits both a high lithium-ion transference number (0.64) and high ionic conductivity (2.03 m S/cm) as well as superior interfacial compacity with lithium *** Li Fe PO_(4)-Li cell using this novel GPE can operate steadily at 2C for 300 cycles,remaining a high discharge capacity of 125 m Ah/g and dendrite-free *** performance improvements for the Li-Li and Cu-Li cells are also presented.

关键词： Polyoxadiazole sulfonate Single-ion conductor Gel polymer electrolyte Lithium dendrite Lithium metal battery

来源：评论

学校读者我要写书评

暂无评论

Photocatalytic degradation of antibiotic sulfamethizole by visible light activated perovskite LaZnO_(3)

引用

Journal of Environmental sciences 2024年第10期144卷 212-224页

作者： Bui The Huy XCuong Nguyen Vu Khac Hoang Bui Nguyen Ngoc Tri Iqra Rabani Nhu Hoa Thi Tran Quang Viet Ly Hai Bang Truong Major of Biomedical Engineering Division of Smart HealthcareCollege of Information Technology and ConvergencePukyong National UniversityBusan 48513Korea Center for Advanced Chemistry Institute of Research and DevelopmentDuy Tan UniversityDa Nang 550000Viet Nam Faculty of Environmental and Chemical Engineering Duy Tan UniversityDa Nang 550000Viet Nam Department of Environment and Energy Sejong UniversitySeoul 05006Korea Lab of Computational Chemistry and Modelling Department of ChemistryFaculty of Natural SciencesQuy Nhon UniversityQuy NhonViet Nam Department of Nanotechnology and Advanced Materials Engineering Sejong UniversitySeoul 05006Korea Faculty of Materials Science and Technology University of ScienceHo Chi Minh City 700000Viet Nam Vietnam National University Ho Chi Minh City 700000Viet Nam Department of Environmental Engineering Seoul National University of Science and TechnologySeoul 01811Korea Optical Materials Research Group Science and Technology Advanced InstituteVan Lang UniversityHo Chi Minh CityViet Nam Faculty of Applied Technology School of TechnologyVan Lang UniversityHo Chi Minh CityViet Nam

In thiswork,the perovskite LaZnO_(3) was synthesized via sol-gel method and applied for photocatalytic treatment of sulfamethizole(SMZ)antibiotics under visible light *** was almost completely degraded(99.2%±0.3%)within 4 hr by photocatalyst LaZnO_(3) at the optimal dosage of 1.1 g/L,with amineralization proportion of 58.7%±0.4%.The efficient performance of LaZnO_(3) can be attributed to itswide-range light absorption and the appropriate energy band edge levels,which facilitate the formation of active agents such as·O_(2)^(−),h^(+),and·*** integration of RP-HPLC/Q-TOF-MS and DFT-based computational techniques revealed three degradation pathways of SMZ,which were initiated by the deamination reaction at the aniline ring,the breakdown of the sulfonamidemoieties,and a process known as Smile-type rearrangement and SO2 *** toxicity of SMZ and the intermediateswere analyzed by quantitative structure activity relationship(QSAR),indicating the effectiveness of LaZnO_(3)-based photocatalysis in preventing secondary pollution of the intermediates to the ecosystem during the degradation *** visible-light-activated photocatalyst LaZnO_(3) exhibited efficient performance in the occurrence of inorganic anions and maintained high durability across multiple recycling tests,making it a promising candidate for practical antibiotic treatment.

关键词： Perovskite Antibiotics Photocatalysis Density functional theory Toxicity

来源：评论

学校读者我要写书评

暂无评论

Machine Learning in Fraud Detection for Financial Sustainability in Credit Card 2nd

Machine Learning in Fraud Detection for Financial Sustainabi...

引用

2nd International Conference on Sustainable Development Using Machine Learning, Artificial Intelligence, and IoT, ICSD 2024

作者： Kaur, Harleen Chauhan, Ritu Shamsi, Yezdanul Haque Alankar, Bhavya Artificial Intelligence and IoT Lab Center for Computational Biology and Bioinformatics Amity University Noida India Department of Computer Science and Engineering School of Engineering Sciences and Technology Jamia Hamdard New Delhi India

ISBN: (纸本)9783031717284

In today’s financial world, various technologies have been emerged and with the evolution in the technologies it has led to the increase in the fraud cases. The sustainable growth of businesses and the financial markets are severely harmed by financial statement fraud. Consequently, it is crucial and vital to build an efficient methodology to identify financial statement fraud in businesses. Usage of credit card are very frequent these days and due to loopholes in the system fraud cases in credit card is rising which causes huge financial losses. In this paper the focus is on credit card fraud detection in real world with the help of machine learning. In this case Logistic Regression, Decision tree method and SVM (support vector machine) are applied on credit card fraud detection for knowledge discovery among the data. The applied methods can detect the fraudulent transactions with highest accuracy. However, the results are predicted using evaluation metrics such as accuracy, error rate, specificity, sensitivity, and ROC curves are calculated to assess each model's performance. Hence, the predicted results suggested that logistic regression is the most accurate model for prediction of fraud datasets in financial sustainability. © The Author(s), under exclusive license to Springer Nature Switzerland AG 2025.

关键词： Decentralized finance

来源：评论

学校读者我要写书评

暂无评论

Liutex-based vortex control with implications for cavitation suppression

引用

Journal of Hydrodynamics 2021年第1期33卷 74-85页

作者： Yi-qian Wang Hai-dong Yu Wei-wen Zhao De-cheng Wan School of Mathematical Science Soochow UniversitySuzhou 215006China School of Aerospace Engineering Tsinghua UniversityBeijing 100084China Computational Marine Hydrodynamics Lab(CMHL) State Key Laboratory of Ocean EngineeringSchool of Naval ArchitectureOcean and Civil EngineeringShanghai Jiao Tong UniversityShanghai 200240China

Viewed as sinews and muscles of fluid motion,coherent vortical structures with their interactions are key to understanding the flow *** upon this observation,we explore the possibility of efficient flow control by directly manipulating vortices numerically inside the flow field based on the vortex definition and identification system of *** objective is twofold:(1)to study the vortex dynamics,for example,by observing the response of the flow to strengthening or weakening of certain vortices,and(2)to obtain efficient vortex-based control strategies which might lead us to practical *** the present numerical study,the manipulating of vortices is achieved by introducing additional source(force)terms to the Navier-Stokes equations,which hereafter will be collectively called Liutex force field *** including controlling the rotation strength and centripetal force of particular vortices are detailed in a flow past a cylinder with different control purposes at Reynolds number of *** examples are provided with a cavitating flow around two-dimensional Clark-Y hydrofoil,with particular interests on cavitation *** is illustrated particular vortex with cavitation encircled could be effectively suppressed.

关键词： Vortex dynamics Liutex vector flow control cavitation Introduction

来源：评论

学校读者我要写书评

暂无评论

Communicative Grounding of Analogical Explanations in Dialogue: A Corpus Study of Conversational Management Acts and Statistical Sequence Models for Tutoring through Analogy

Communicative Grounding of Analogical Explanations in Dialog...

引用

2021 Conference on Reasoning and Interaction, ReInAct 2021

作者： Del-Bosque-Trevino, Jorge Hough, Julian Purver, Matthew Education Lab. Computational Linguistics Lab. Cognitive Science Research Group School of Electronic Engineering and Computer Science Queen Mary University of London United Kingdom Human Interaction Lab. Computational Linguistics Lab. Cognitive Science Research Group School of Electronic Engineering and Computer Science Queen Mary University of London United Kingdom Jožef Stefan Institute Computational Linguistics Lab. Cognitive Science Research Group School of Electronic Engineering and Computer Science Queen Mary University of London United Kingdom

ISBN: (纸本)9781955917070

We present a conversational management act (CMA) annotation schema for one-to-one tutorial dialogue sessions where a tutor uses an analogy to teach a student a concept. CMAs are more fine-grained sub-utterance acts compared to traditional dialogue act markup. The schema achieves an inter-annotator agreement (IAA) Cohen Kappa score of at least 0.66 across all 10 classes. We annotate a corpus of analogical episodes with the schema and develop statistical sequence models from the corpus which predict tutor content related decisions, in terms of the selection of the analogical component (AC) and tutor conversational management act (TCMA) to deploy at the current utterance, given the student’s behaviour. CRF sequence classifiers perform well on AC selection and robustly on TCMA selection, achieving respective accuracies of 61.9% and 56.3% on a cross-validation experiment over the corpus. © 2021 Association for computational Linguistics.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Embracing Safe Contacts with Contact-aware Planning and Control

arXiv

引用

arXiv 2023年

作者： Li, Zhaoting Zamora, Miguel Zheng, Hehui Coros, Stelian Computational Robotics Lab Computer Science Department ETH Zurich Switzerland Soft Robotics Lab Department of Mechanical and Process Engineering ETH Zurich Switzerland

Unlike human beings that can employ the entire surface of their limbs as a means to establish contact with their environment, robots are typically programmed to interact with their environments via their end-effectors, in a collision-free fashion, to avoid damaging their environment. In a departure from such a traditional approach, this work presents a contact-aware controller for reference tracking that maintains interaction forces on the surface of the robot below a safety threshold in the presence of both rigid and soft contacts. Furthermore, we leveraged the proposed controller to extend the BiTRRT sample-based planning method to be contact-aware, using a simplified contact model. The effectiveness of our framework is demonstrated in hardware experiments using a Franka robot in a setup inspired by the Amazon stowing task. A demo video of our results can be seen here: https://***/2WeYytauhNg. © 2023, CC BY.

关键词： End effectors

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：