检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Hoshino, Hikaru Li, Jiaxing Menon, Arnav Dolan, John M. Nakahira, Yorie The Department of Electrical Materials and Engineering University of Hyoto Japan The Department of Electrical and Computer Engineering Carnegie Mellon University United States The Robotics Institute Carnegie Mellon University United States

This paper proposes a novel learning-based framework for autonomous driving based on the concept of maximal safety probability. Efficient learning requires rewards that are informative of desirable/undesirable states, but such rewards are challenging to design manually due to the difficulty of differentiating better states among many safe states. On the other hand, learning policies that maximize safety probability does not require laborious reward shaping but is numerically challenging because the algorithms must optimize policies based on binary rewards sparse in time. Here, we show that physics-informed reinforcement learning can efficiently learn this form of maximally safe policy. Unlike existing drift control methods, our approach does not require a specific reference trajectory or complex reward shaping, and can learn safe behaviors only from sparse binary rewards. This is enabled by the use of the physics loss that plays an analogous role to reward shaping. The effectiveness of the proposed approach is demonstrated through lane keeping in a normal cornering scenario and safe drifting in a high-speed racing scenario. © 2024, CC BY.

关键词： Contrastive Learning

来源：评论

学校读者我要写书评

暂无评论

Codes for Correcting t Limited-Magnitude Sticky Deletions

Codes for Correcting t Limited-Magnitude Sticky Deletions

引用

2023 IEEE International Symposium on Information Theory, ISIT 2023

作者： Wang, Shuche Vu, Van Khu Tan, Vincent Y. F. National University of Singapore Institute of Operations Research and Analytics Singapore National University of Singapore Department of Mathematics Singapore National University of Singapore Department of Electrical and Computer Engineering Singapore National University of Singapore Department of Industrial Systems Engineering and Management Singapore

ISBN: (纸本)9781665475549

Codes for correcting sticky insertions/deletions and limited-magnitude errors have attracted significant attention due to their applications of flash memories, racetrack memories, and DNA data storage systems. In this paper, we first consider the error type of t sticky deletions with l-limited-magnitude and propose a non-systematic code for correcting this type of error with redundancy 2t(1 - 1/p) • log(n + 1) + O(1), where p is the smallest prime larger than l + 1. Next, we present a systematic code construction with an efficient encoding and decoding algorithm with redundancy ? 2t(1 - 1/p)}? · ? log p?/ log p log (n + 1) + O(log log n), where p is the smallest prime larger than l + 1. © 2023 IEEE.

关键词： Redundancy

来源：评论

学校读者我要写书评

暂无评论

Principled penalty-based methods for bilevel reinforcement learning and RLHF 24

Principled penalty-based methods for bilevel reinforcement l...

引用

Proceedings of the 41st International Conference on Machine Learning

作者： Han Shen Zhuoran Yang Tianyi Chen Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute Department of Statistics and Data Science Yale University

Bilevel optimization has been recently applied to many machine learning tasks. However, their applications have been restricted to the supervised learning setting, where static objective functions with benign structures are considered. But bilevel problems such as incentive design, inverse reinforcement learning (RL), and RL from human feedback (RLHF) are often modeled as dynamic objective functions that go beyond the simple static objective structures, which pose significant challenges of using existing bilevel solutions. To tackle this new class of bilevel problems, we introduce the first principled algorithmic framework for solving bilevel RL problems through the lens of penalty formulation. We provide theoretical studies of the problem landscape and its penalty-based (policy) gradient algorithms. We demonstrate the effectiveness of our algorithms via simulations in the Stackelberg game and RLHF.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Learning-Based Joint Resource Slicing and Scheduling in Space-Terrestrial Integrated Vehicular Networks

引用

Journal of Communications and Information Networks 2021年第3期6卷 208-223页

作者： Huaqing Wu Jiayin Chen Conghao Zhou Junling Li Xuemin(Sherman)Shen Department of Electrical and Computer Engineering University of WaterlooWaterloo.ON N2L 3G1Canada Department of Electrical and Computer Engineering the University of British ColumbiaVancouverBC V6T 1Z4Canada Shenzhen Institute of Artificial Intelligence and Robotics for Society(AIRS) and the Department of Electrical and Computer EngineeringUniversity of WaterlooCanada

In this paper,we investigate the resource slicing and scheduling problem in the space-terrestrial integrated vehicular networks to support both delay-sensitive services(DSSs)and delay-tolerant services(DTSs).Resource slicing and scheduling are to allocate spectrum resources to different slices and determine user association and bandwidth allocation for individual *** accommodate the dynamic network conditions,we first formulate a joint resource slicing and scheduling(JRSS)problem to minimize the long-term system cost,including the DSS requirement violation cost,DTS delay cost,and slice reconfiguration *** resource slicing and scheduling decisions are interdependent with different timescales,we decompose the JRSS problem into a large-timescale resource slicing subproblem and a small-timescale resource scheduling *** propose a two-layered reinforcement learning(RL)-based JRSS scheme to find the solutions to the *** the resource slicing layer,spectrum resources are pre-allocated to different slices via a proximal policy optimization-based RL *** the resource scheduling layer,spectrum resources in each slice are scheduled to individual vehicles based on dynamic network conditions and service requirements via matching-based *** conduct extensive trace-driven experiments to demonstrate that the proposed scheme can effectively reduce the system cost while satisfying service quality requirements.

关键词： space-terrestrial integrated vehicular networks LEO satellite communication resource slicing and scheduling reinforcement learning matching-based optimization

来源：评论

学校读者我要写书评

暂无评论

Exploring 3D Reconstruction with Drone Images: Advances and Challenges in Urban Environments 50

Exploring 3D Reconstruction with Drone Images: Advances and ...

引用

50th Annual Conference of the IEEE Industrial Electronics Society, IECON 2024

作者： Miranda Baldivieso, Thiago João da Silva Batista, Taise Grazielle Chagas, Fabio Suim Rodrigues Velho, Luiz Carlos Pacheco Ferreira Rosa, Paulo Fernando Defense Engineering Graduate Program Military Institute of Engineering Rio de Janeiro Brazil State Institute of the Environment Inea RJ Rio de Janeiro Brazil Department of Science and Industry systems University of South-Eastern Norway Kongsberg Norway Department of Computer Graphics Institute for Pure and Applied Mathematics Rio de Janeiro Brazil Electrical Engineering Graduate Program Military Institute of Engineering Rio de Janeiro Brazil

ISBN: (纸本)9781665464543

This working-in-progress paper aims to present a three-dimensional reconstruction using aerial images in different environments. The experiments were conducted with aircraft in both external and internal settings, starting with image acquisition, followed by the application of specific photogrammetry software-both commercial and open-source-and concluding with a qualitative evaluation of the results. © 2024 IEEE.

关键词： Aerial photography

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Communication Protocols for IoT Devices 9

Energy-Efficient Communication Protocols for IoT Devices

引用

9th International Conference on Science, Technology, engineering and Mathematics, ICONSTEM 2024

作者： Rahamathunnisa, U. Vivekanand, Chettiyar Vani Chithras, T. Sreedevi, E. Kiran, Venneti Rajendiran, M. School of Computer Science Engineering And Information Systems Vellore Institute of Technology Tamil Nadu Vellore632014 India Rmk College of Engineering And Technology Department of Electronics And Communication Engineering Tamil Nadu Thiruvallur India Veltech Rangarajan Dr.Sagunthala R&d Institute of Science And Technology Department of Electrical And Electronics Engineering Chennai Tamil Nadu Avadi600062 India KoneruLakshmaiah Education Foundation Department of Computer Science And Engineering Andhra Pradesh Vaddeswaram India Andhra Pradesh Surampalem India Panimalar Engineering College Department of Computer Science And Engineering Tamil Nadu Chennai600123 India

ISBN: (纸本)9798350365092

The development and optimisation of energy-efficient communication protocols that are specifically adapted for Internet of Things (IoT) devices is the focus of this research. The proliferation of interoperable bias has given rise to the converse girding energy consumption, which has come a consummate concern in the contemporary period. The primary end of this bid is to probe innovative approaches and protocols with the capacity to reduce the energy consumption of Internet of Effects(IoT) connections. As a result, the bias will retain the capability to attain an extended functional lifetime while coincidently producing a lowered environmental footmark. The presented protocols aim to achieve a result that balances the utilisation of energy coffers with the conservation of dependable communication. A thorough examination of the presently employed protocols, coupled with the perpetration of sophisticated optimisation ways, will be the styles employed to achieve this thing. Theprovides a significant donation to the development and prosecution of communication protocols that effectively address the unique energy constraints encountered by Internet of effects(10T)- connected ecosystems. This contributes to the enhancement of effectiveness and sustainability in the world of networked bias, which is expanding at an accelerated rate. © 2024 IEEE.

关键词： Energy efficiency

来源：评论

学校读者我要写书评

暂无评论

Shared Information Under Simple Markov Independencies

Shared Information Under Simple Markov Independencies

引用

IEEE International Symposium on Information Theory

作者： Madhura Pathegama Sagnik Bhattacharya Department of Electrical and Computer Engineering Institute for Systems Research University of Maryland College Park MD USA

ISBN: (数字)9798350382846

ISBN: (纸本)9798350382853

Shared information is a measure of mutual dependence among $m\geq 2$ jointly distributed discrete random variables. We show that the shared information of a Markov random field in which the underlying graph has at least one cut vertex, is the same as the minimum shared information of its blocks (also called biconnected components). This generalizes prior results on shared information of Markov random fields to a much wider class of nontree graphs.

关键词： Random variables Markov random fields Information theory

来源：评论

学校读者我要写书评

暂无评论

Local Fading Memory Effects in a Tantalum Oxide ReRAM Cell from Hewlett Packard Labs 2

Local Fading Memory Effects in a Tantalum Oxide ReRAM Cell f...

引用

2nd Edition IEEE International Conference on Metrology for eXtended Reality, Artificial Intelligence and Neural engineering, MetroXRAINE 2023

作者： Ascoli, A. Schmitt, N. Messaris, I. Demirkol, A.S. Tetzlaff, R. Strachan, J.P. Chua, L.O. Institute of Circuits and Systems Technische Universität Dresden Faculty of Electrical and Computer Engineering Dresden Germany Peter Grünberg Institute Forschungszentrum Jülich GmbH Jülich Germany Rwth Aachen University Aachen Germany University of California Berkeley Department of Electrical Engineering and Computer Sciences BerkeleyCA94720 United States

ISBN: (纸本)9798350300802

Memristors are nano-devices, which, in some of their physical realisations, feature multiple operating modes, allowing data to be alternatively sensed, stored, or processed. Their multi-functionality, which resembles the operating principles of neural structures in the brain, in conjunction with the ease of their fabrication, which is typically carried out on top of underlying CMOS circuitry, at the regularly-spaced spaced crossings of crossbar matrices, composed of two vertically-adjacent and perpendicular sets of parallel metal lines, makes their use in integrated circuit design ideally suitable for enhancing the performance of state-of-the-art computing machines without the need to shrink the sizes of CMOS transistors any further. Given the extraordinary opportunities, which memristors open up in the field of Artificial Intelligence, gaining a deep understanding of their inherently-nonlinear dynamics is of primary importance. Fading memory is the capability of a system to display a unique steady state under a stimulus from a given class. Fading memory effects in a memristive medium were first discovered in 2016 through a comprehensive experimental and theoretical investigation of the switching dynamics of a TaOx ReRAM cell. Recently, a bifurcation study of a predictive model of the same nanostructure revealed how it may also exhibit bi-stability under suitable AC periodic excitation. Bi-stability is essentially a local form of fading memory. This paper gains a deeper insight into the local fading memory properties of the TaOx ReRAM cell, demonstrating how the existence domain for its memory state may be in fact partitioned in a tuneable number of basins of attraction, each of which includes all the initial conditions, from which the memristor would settle into one of the admissible locally-stable oscillatory steady states. © 2023 IEEE.

关键词： Fading Memory Local Fading Memory Memory Memristor Multi-stability Non-volatility ReRAM

来源：评论

学校读者我要写书评

暂无评论

Earthworm Optimization with Improved SqueezeNet Enabled Facial Expression Recognition Model

引用

computer systems Science & engineering 2023年第8期46卷 2247-2262页

作者： N.Sharmili Saud Yonbawi Sultan Alahmari E.Laxmi Lydia Mohamad Khairi Ishak Hend Khalid Alkahtani Ayman Aljarbouh Samih M.Mostafa Computer Science and Engineering Department Gayatri Vidya Parishad College of Engineering for WomenVisakhapatnamAndhra PradeshIndia Department of Software Engineering College of Computer Science and EngineeringUniversity of JeddahJeddahSaudi Arabia King Abdul Aziz City for Science and Technology RiyadhKingdom of Saudi Arabia Department of Computer Science and Engineering Vignan’s Institute of Information TechnologyVisakhapatnam530049India School of Electrical and Electronic Engineering Engineering CampusUniversiti Sains Malaysia(USM)Nibong TebalPenang14300Malaysia Department of Information Systems College of Computer and Information SciencesPrincess Nourah bint Abdulrahman UniversityRiyadh11564Saudi Arabia Department of Computer Science University of Central AsiaNaryn722600Kyrgyzstan Faculty of Computers and Information South Valley UniversityQena83523Egypt

Facial expression recognition(FER)remains a hot research area among computer vision researchers and still becomes a challenge because of high intraclass *** techniques for this problem depend on hand-crafted features,namely,LBP,SIFT,and HOG,along with that a classifier trained on a database of videos or *** execute perform well on image datasets captured in a controlled condition;however not perform well in the more challenging dataset,which has partial faces and image ***,many studies presented an endwise structure for facial expression recognition by utilizing DL ***,this study develops an earthworm optimization with an improved SqueezeNet-based FER(EWOISN-FER)*** presented EWOISN-FER model primarily applies the contrast-limited adaptive histogram equalization(CLAHE)technique as a pre-processing *** addition,the improved SqueezeNet model is exploited to derive an optimal set of feature vectors,and the hyperparameter tuning process is performed by the stochastic gradient boosting(SGB)***,EWO with sparse autoencoder(SAE)is employed for the FER process,and the EWO algorithm appropriately chooses the SAE ***-ranging experimental analysis is carried out to examine the performance of the proposed *** experimental outcomes indicate the supremacy of the presented EWOISN-FER technique.

关键词： Facial expression recognition deep learning computer vision earthworm optimization hyperparameter optimization

来源：评论

学校读者我要写书评

暂无评论

Expeditious Diagnosis of Cerebral Aneurysm on MR Brain Angiogram Images Utilizing Soft Computing Techniques 7

Expeditious Diagnosis of Cerebral Aneurysm on MR Brain Angio...

引用

7th International Conference on Inventive Computation Technologies, ICICT 2024

作者： Anjaneya, L.H. Maheswari, K. Revathi, N. Siva, K.P. Soudagar, Manzoore Elahi M. Dhanraj, Joshuva Arockia Bapuji Institute of Engineering And Technology Department of Electrical And Electronics Engineering Karnataka Davangere577004 India Cmr Technical Campus Department of Computer Science And Engineering Hyderabad Telangana Kandlakoya501401 India Panimalar Engineering College Department of Electronics And Communication Engineering Tamil Nadu Chennai600123 India Sri Eshwar College of Engineering Department of Computer Science And Business Systems Tamil Nadu Coimbatore641202 India Saveetha Institute of Medical And Techinical Sciences Saveetha School of Engineering Department of Vlsi Microelectronics Tamil Nadu Chennai602105 India Department of Mechatronics Engineering Chennai Tamil Nadu Padur603103 India

ISBN: (纸本)9798350359299

It is always a challenging task for radiologists to detect Cerebral Aneurysms (CA). Unenhanced magnetic resonance angiography (MRA) has become a popular screening tool because of its non-invasiveness, lack of contrast, and ionizing radiation. Because of its screening role, MRA requires a high recognition rate. This paper presents Machine Learning (ML) methods for detecting CA in MRA data. MRA scans of normal and aneurysms are collected and evaluated. This research study is divided into two sections: segmentation and classification. First, K-Means Clustering (KMC) is used for aneurysm segmentation from MR images. Second, image classification is done. For that, the appropriate features are retrieved using feature extraction and selection. Then Gray Level Run Length Matrix (GLRLM) is used for feature extraction and the Genetic Algorithm (GA) for feature selection. The GLRLM and GA features are provided independently to the ML models for training and testing. The ML models used in the research work are Artificial Neural Networks (ANN) and Modified Fuzzy C-Means (MFCM). The metrics are employed to evaluate the GLRM and GA features with ANN and MFCM. The combination of GA features with ANN produces the best results when compared to other combinations. The GA with ANN obtains an outstanding 98.33% accuracy, 100% specificity and precision, 96.67% sensitivity, and 98.31% F1 score. © 2024 IEEE.

关键词： Genetic algorithms

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：