检索结果-内蒙古大学图书馆

arXiv 2023年

作者： Dwivedi, Anmol Tajer, Ali Electrical Computer and Systems Engineering Department Rensselaer Polytechnic Institute United States

This paper proposes a data-driven graphical framework for the real-time search of risky cascading fault chains (FCs). While identifying risky FCs is pivotal to alleviating cascading failures, the complex spatio-temporal dependencies among the components of the power system render challenges to modeling and analyzing FCs. Furthermore, the real-time search of risky FCs faces an inherent combinatorial complexity that grows exponentially with the size of the system. The proposed framework leverages the recent advances in graph recurrent neural networks to circumvent the computational complexities of the real-time search of FCs. The search process is formalized as a partially observable Markov decision process (POMDP), which is subsequently solved via a time-varying graph recurrent neural network (GRNN) that judiciously accounts for the inherent temporal and spatial structures of the data generated by the system. The key features of this structure include (i) leveraging the spatial structure of the data induced by the system topology, (ii) leveraging the temporal structure of data induced by system dynamics, and (iii) efficiently summarizing the system’s history in the latent space of the GRNN. The proposed framework’s efficiency is compared to the relevant literature on the IEEE 39-bus New England system and the IEEE 118-bus system. © 2023, CC BY.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Experimental demonstration of SnO_(2) nanofiber-based memristors and their data-driven modeling for nanoelectronic applications

引用

Chip 2023年第4期2卷 142-153页

作者： Soumi Saha Madadi Chetan Kodand Reddy Tati Sai Nikhil Kaushik Burugupally Sanghamitra DebRoy Akshay Salimath Venkat Mattela Surya Shankar Dan Parikshit Sahatiya Department of Electrical&Electronics Engineering Birla Institute of Technology and Science PilaniHyderabad CampusHyderabad 500078India Department of Computer Science and Information Systems Birla Institute of Technology and Science PilaniHyderabad CampusHyderabad 500078India Nanomagnetic Division Ceremorphic India Pvt LtdIndia Ceremorphic Inc.San JoseCA 95131USA

This paper demonstrated the fabrication,characterization,datadriven modeling,and practical application of a 1D SnO_(2)nanofiber-based memristor,in which a 1D SnO_(2)active layer wassandwiched between silver(Ag)and aluminum(Al)*** yielded a very high ROFF:RON of~104(ION:IOFF of~105)with an excellent activation slope of 10 mV/dec,low set voltage ofVSET~1.14 V and good *** paper physically explained the conduction mechanism in the layered SnO_(2)*** conductive network was composed of nanofibersthat play a vital role in the memristive action,since more conductive paths could facilitate the hopping of electron *** structures experimentally extracted with the adoption of ultraviolet photoelectron spectroscopy strongly support the claimsreported in this *** machine learning(ML)–assisted,datadriven model of the fabricated memristor was also developedemploying different popular algorithms such as polynomialregression,support vector regression,k nearest neighbors,andartificial neural network(ANN)to model the data of the *** have proposed two types of ANN models(type I andtype II)algorithms,illustrated with a detailed flowchart,to modelthe fabricated *** with standard ML techniques shows that the type II ANN algorithm provides the bestmean absolute percentage error of 0.0175 with a 98%R^(2)*** proposed data-driven model was further validated with the characterization results of similar new memristors fabricated adoptingthe same fabrication recipe,which gave satisfactory ***,the ANN type II model was applied to design and implementsimple AND&OR logic functionalities adopting the fabricatedmemristors with expected,near-ideal characteristics.

关键词： Nanofiber-based memristors Data-driven modeling Artificial neural network(ANN) SnO_(2)

来源：评论

学校读者我要写书评

暂无评论

MCTS-Judge: Test-Time Scaling in LLM-as-a-Judge for Code Correctness Evaluation

arXiv

引用

arXiv 2025年

作者： Wang, Yutong Ji, Pengliang Yang, Chaoqun Li, Kaixin Hu, Ming Li, Jiaoyang Sartoretti, Guillaume Department of Mechanical Engineering National University of Singapore Singapore Robotics Institute Carnegie Mellon University United States School of Computing National University of Singapore Singapore Department of Electrical and Computer Systems Monash University Australia

The LLM-as-a-Judge paradigm shows promise for evaluating generative content but lacks reliability in reasoning-intensive scenarios, such as programming. Inspired by recent advances in reasoning models and shifts in scaling laws, we pioneer bringing test-time computation into LLM-as-a-Judge, proposing MCTS-Judge, a resource-efficient, System-2 thinking framework for code correctness evaluation. MCTS-Judge leverages Monte Carlo Tree Search (MCTS) to decompose problems into simpler, multi-perspective evaluations. Through a node-selection strategy that combines self-assessment based on historical actions in the current trajectory and the Upper Confidence Bound for Trees based on prior rollouts, MCTS-Judge balances global optimization and refinement of the current trajectory. We further designed a high-precision, unit-test-level reward mechanism to encourage the Large Language Model (LLM) to perform line-by-line analysis. Extensive experiments on three benchmarks and five LLMs demonstrate the effectiveness of MCTS-Judge, which improves the base model’s accuracy from 41.0% to 80.0%, surpassing the o1-series models with 3× fewer tokens. Further evaluations validate the superiority of its reasoning trajectory in logic, analytics, thoroughness, and overall quality, while revealing the test-time scaling law of the LLM-as-a-Judge paradigm. Copyright © 2025, The Authors. All rights reserved.

关键词： Scaling laws

来源：评论

学校读者我要写书评

暂无评论

On Penalty-based Bilevel Gradient Descent Method

arXiv

引用

arXiv 2023年

作者： Shen, Han Xiao, Quan Chen, Tianyi Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute United States

Bilevel optimization enjoys a wide range of applications in emerging machine learning and signal processing problems such as hyper-parameter optimization, image reconstruction, meta-learning, adversarial training, and reinforcement learning. However, bilevel optimization problems are traditionally known to be difficult to solve. Recent progress on bilevel algorithms mainly focuses on bilevel optimization problems through the lens of the implicit-gradient method, where the lower-level objective is either strongly convex or unconstrained. In this work, we tackle a challenging class of bilevel problems through the lens of the penalty method. We show that under certain conditions, the penalty reformulation recovers the (local) solutions of the original bilevel problem. Further, we propose the penalty-based bilevel gradient descent (PBGD) algorithm and establish its finite-time convergence for the constrained bilevel problem with lower-level constraints yet without lower-level strong convexity. Experiments on synthetic and real datasets showcase the efficiency of the proposed PBGD algorithm. The code for implementing this algorithm is publicly available on GitHub. Copyright © 2023, The Authors. All rights reserved.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Design of Capsule Robot for Magnetic Gastric Biopsy and Drug Administration 20

Design of Capsule Robot for Magnetic Gastric Biopsy and Drug...

引用

20th IEEE International Conference on Mechatronics and Automation, ICMA 2023

作者： Xu, Zhihui Li, Zongze Liu, Shoubin Hu, Chengzhi Shenzhen Key Laboratory of Biomimetic Robotics and Intelligent Systems Department of Mechanical and Energy Engineering Southern University of Science and Technology Shenzhen China Guangdong Prov. Key Laboratory of Human-Augmentation and Rehabilitation Robotics in Universities Southern University of Science and Technology Shenzhen China Harbin Institute of Technology School of Mechanical and Electrical Engineering Shenzhen China

ISBN: (纸本)9798350320831

Capsule robots have emerged as a promising tool for the early diagnosis of gastrointestinal diseases. Unlike conventional gastroscopy, they offer the potential for minimally invasive or even non-invasive treatment, and can alleviate discomfort and pain in patients. However, current capsule robots have some limitations. For instance, they lack stomach tissue biopsy and drug administration functions, and the capsule robot used for biopsy does not consider the tolerance of gastrointestinal tissue. Similarly, the capsule robot used for drug administration does not account for the outlet speed of drug administration. This paper proposes a new biopsy method using a magnetically manipulated capsule robot and analyzes the principles governing the robot's manipulation during the biopsy and drug administration processes. Furthermore, an ANSYS gastric model was established and used to simulate and experimentally analyze damage to the gastric tissue model from different types of biopsy needles and their moving speeds. The appropriate type of biopsy needle and moving speed were determined to prevent damage to the gastric tissue. The study also used FLUENT to determine the appropriate thrust at the entrance of the drug cavity and the impact of the drug density on the drug outlet velocity to ensure complete adherence of the drug to the lesion. © 2023 IEEE.

关键词： Biopsy

来源：评论

学校读者我要写书评

暂无评论

Movement-Based Interaction: An Approach to Visual and Auditory Stimulus and Memory*

Movement-Based Interaction: An Approach to Visual and Audito...

引用

Experiment@ International Conference (exp.at)

作者： Maria Rita Nogueira Paulo Menezes Institute of Systems and Robotics University of Coimbra Coimbra Portugal Dept. of Electrical and Computer Engineering University of Coimbra Coimbra Portugal

ISBN: (数字)9798350314403

ISBN: (纸本)9798350314410

Movement-based interaction design is an innovative field that leverages the body's natural movements to create intuitive and engaging interfaces. This approach takes a sensorial approach to visual and auditory memory, seeking to enhance user engagement and create more meaningful experiences. The use of movement-based interfaces has significant implications for senior citizens, who may have mobility limitations or cognitive impairments. Using movement-based interfaces, seniors can engage in physical activity and maintain cognitive function, essential aspects for healthy aging. Movement-based interfaces can also help children develop physical coordination, cognitive skills, and social interaction. Furthermore, movement-based interaction design can benefit everyone's physical well-being by promoting physical activity and reducing sedentary behavior. Studies have demonstrated that movement-based interfaces can enhance posture, physical activity levels, and general health. By combining movement with sensory experiences, movement-based interaction design can create a more engaging and enjoyable experience for users, which can motivate them to maintain physical and cognitive health. In this article, we present two case studies that research the effectiveness of movement-based interaction design in enhancing visual and auditory memory. Overall, the results of these case studies suggest that movement-based interaction design is an effective approach to enhancing visual and auditory memory. The introduction of movement-based interfaces can give participants a more enjoyable and memorable experience, improving user engagement and information retention.

关键词： Visualization Pediatrics Aging Older adults

来源：评论

学校读者我要写书评

暂无评论

Sample complexity of neural policy mirror descent for policy optimization on low-dimensional manifolds

The Journal of Machine Learning Research

引用

The Journal of Machine Learning Research 2024年第1期25卷 11019-11085页

作者： Zhenghao Xu Xiang Ji Minshuo Chen Mengdi Wang Tuo Zhao H. Milton Stewart School of Industrial and Systems Engineering Georgia Institute of Technology Atlanta GA Department of Electrical and Computer Engineering Princeton University Princeton NJ

Policy gradient methods equipped with deep neural networks have achieved great success in solving high-dimensional reinforcement learning (RL) problems. However, current analyses cannot explain why they are resistant to the curse of dimensionality. In this work, we study the sample complexity of the neural policy mirror descent (NPMD) algorithm with deep convolutional neural networks (CNN). Motivated by the empirical observation that many high-dimensional environments have state spaces possessing low-dimensional structures, such as those taking images as states, we consider the state space to be a d-dimensional manifold embedded in the D-dimensional Euclidean space with intrinsic dimension d ≪ D. We show that in each iteration of NPMD, both the value function and the policy can be well approximated by CNNs. The approximation errors are controlled by the size of the networks, and the smoothness of the previous networks can be inherited. As a result, by properly choosing the network size and hyperparameters, NPMD can find an ε-optimal policy with Õ(ε-d/α-2) samples in expectation, where α ∈ (0, 1] indicates the smoothness of environment. Compared to previous work, our result exhibits that NPMD can leverage the low-dimensional structure of state space to escape from the curse of dimensionality, explaining the efficacy of deep policy gradient algorithms.

关键词： deep reinforcement learning policy optimization function approximation convolutional neural network Riemannian manifold

来源：评论

学校读者我要写书评

暂无评论

Sudden Cardiac Arrest Detection Based on Temporal Features of ECG Using Support Vector Machine Classifier

Sudden Cardiac Arrest Detection Based on Temporal Features o...

引用

International Conference on Industrial Instrumentation and Control,ICI2C 2021

作者： Banerjee, Prakash Bhattacherjee, Saptak Dasgupta, Kousik Department of Electrical and Electronics Engineering University of Engineering and Management Kolkata India School of Mechatronics and Robotics Indian Institute of Engineering Science and Technology Howrah Shibpur India Department of Computer Science and Engineering Kalyani Government Engineering College Kalyani India

ISBN: (纸本)9789811670107

Sudden cardiac arrest (SCA) is caused due to malfunctioning in heart rhythm, and the patient needs to be hospitalized as early as possible just after the onset of SCA;otherwise, it can lead to the death of the patient. Therefore, detection of SCA with a very high accuracy and within a very short time after the onset of symptoms is very crucial for patient survival. Several approaches for detection of SCA have been made using temporal and morphological variations in the electrocardiogram (ECG). However, the existing methods suffer from low accuracy or involve long-term monitoring of ECG. Therefore, in the present work, an intelligent algorithm to detect SCA based on the extraction of temporal features like RR interval and the QT interval from ECG and subsequently classifying them using support vector machine (SVM) has been proposed. Experimental evaluation of the proposed method on five SCA and five normal subjects demonstrates a high classification accuracy of 100% which proves the superiority of the proposed algorithm as an effective method for detection of SCA. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.

关键词： Electrocardiograms

来源：评论

学校读者我要写书评

暂无评论

Fine-grained Intent Classification in the Legal Domain

Fine-grained Intent Classification in the Legal Domain

引用

2022 Workshop on Scientific Document Understanding, SDU 2022

作者： Mullick, Ankan Nandy, Abhilash Kapadnis, Manav Nitin Patnaik, Sohan Raghav, R. Department of Computer Science and Engineering Indian Institute of Technology Kharagpur India Department of Electrical Engineering Indian Institute of Technology Kharagpur India Department of Mechanical Engineering Indian Institute of Technology Kharagpur India Department of Industrial and Systems Engineering Indian Institute of Technology Kharagpur India

A law practitioner has to go through a lot of long legal case proceedings. To understand the motivation behind the actions of different parties/individuals in a legal case, it is essential that the parts of the document that express an intent corresponding to the case be clearly understood. In this paper, we introduce a dataset of 93 legal documents, belonging to the case categories of either Murder, Land Dispute, Robbery, or Corruption, where phrases expressing intent same as the category of the document are annotated. Also, we annotate fine-grained intents for each such phrase to enable a deeper understanding of the case for a reader. Finally, we analyze the performance of several transformer-based models in automating the process of extracting intent phrases (both at a coarse and a fine-grained level), and classifying a document into one of the possible 4 categories, and observe that, our dataset is challenging, especially in the case of fine-grained intent classification. © 2021 Copyright for this paper by its authors.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

Robust Data-driven Sparse Estimation of Power Flow Sensitivities for Smart Grid Monitoring and Operation 11

Robust Data-driven Sparse Estimation of Power Flow Sensitivi...

引用

11th International Conference on Innovative Smart Grid Technologies - Asia, ISGT-Asia 2022

作者： Liang, Yingqi Zhao, Junbo Kumar, Dhivya Sampath Siano, Pierluigi Srinivasan, Dipti National University of Singapore Department of Electrical and Computer Engineering Singapore Singapore University of Connecticut Department of Electrical and Computer Engineering Storrs United States Singapore Institute of Technology Engineering Cluster Singapore Singapore University of Salerno Department of Management & Innovation Systems SA Fisciano Italy

ISBN: (纸本)9798350399660

Distribution factors (DFs) are sensitivities of active power line flows with respect to system control variables, which play a crucial role in smart grid monitoring and secure operation. This paper proposes a novel data-driven framework for robust sparse DF estimation that can handle the collinearity of high-dimensional synchrophasor measurements in large-scale smart grids. It does not require power flow models, thereby facilitating power flow sensitivity analysis with adaptiveness to operating-point changes and data uncertainties from renewable generations. An Elastic-net (Enet) estimator is used to sparsify DFs, and a new LARS-Enet algorithm is designed to achieve the data-driven estimation. Owing to the l1,2 regularization in Enet, this framework can yield the sparse estimator with robustness to collinearity. Due to the early-stopping property and data-driven settings of LARS-Enet, this framework can solve the estimator fast online. Results carried out on the IEEE 300-bus system indicate the potential of the proposed framework for implementing new grid codes associated with power flow sensitivity analysis and advancing the future development of power-flow security-constrained operations in large-scale smart grids. © 2022 IEEE.

关键词： Sensitivity analysis

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：