检索结果-内蒙古大学图书馆

arXiv 2024年

作者： Zhang, Yixuan Li, Zhidong Wang, Yang Chen, Fang Fan, Xuhui Zhou, Feng School of Statistics and Data Science Southeast University China Data Science Institute University of Technology Sydney Australia School of Computing Macquarie University Australia Center for Applied Statistics School of Statistics Renmin University of China China Beijing Advanced Innovation Center for Future Blockchain and Privacy Computing China

Machine learning algorithms often struggle to eliminate inherent data biases, particularly those arising from unreliable labels, which poses a significant challenge in ensuring fairness. Existing fairness techniques that address label bias typically involve modifying models and intervening in the training process, but these lack flexibility for large-scale datasets. To address this limitation, we introduce a data selection method designed to efficiently and flexibly mitigate label bias, tailored to more practical needs. Our approach utilizes a zero-shot predictor as a proxy model that simulates training on a clean holdout set. This strategy, supported by peer predictions, ensures the fairness of the proxy model and eliminates the need for an additional holdout set, which is a common requirement in previous methods. Without altering the classifier’s architecture, our modality-agnostic method effectively selects appropriate training data and has proven efficient and effective in handling label bias and improving fairness across diverse datasets in experimental evaluations. Copyright © 2024, 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

Controlling the False Discovery Rate in Subspace Selection

arXiv

引用

arXiv 2024年

作者： Díaz, Mateo Chandrasekaran, Venkat Department of Applied Mathematics and Statistics Johns Hopkins University BaltimoreMD21218 United States Departments of Computing and Mathematical Sciences and of Electrical Engineering California Institute of Technology PasadenaCA91125 United States

Controlling the false discovery rate (FDR) is a popular approach to multiple testing, variable selection, and related problems of simultaneous inference. In many contemporary applications, models are not specified by discrete variables, which necessitates a broadening of the scope of the FDR control paradigm. Motivated by the ubiquity of low-rank models for high-dimensional matrices, we present methods for subspace selection in principal components analysis that provide control on a geometric analog of FDR that is adapted to subspace selection. Our methods crucially rely on recently-developed tools from random matrix theory, in particular on a characterization of the limiting behavior of eigenvectors and the gaps between successive eigenvalues of large random matrices. Our procedure is parameter-free, and we show that it provides FDR control in subspace selection for common noise models considered in the literature. We demonstrate the utility of our algorithm with numerical experiments on synthetic data and on problems arising in single-cell RNA sequencing and hyperspectral *** Codes 62H15, 62H25, 62H12, 62R07 Copyright © 2024, The Authors. All rights reserved.

关键词： Principal component analysis

来源：评论

学校读者我要写书评

暂无评论

Fog computing in Optical Access Networks: An Energy-efficient and Deadline-aware Task Scheduling Mechanism

Fog Computing in Optical Access Networks: An Energy-efficien...

引用

International Conference on Information technology in Asia (CITA)

作者： Fathiyah Haji Ahmad S. H. Shah Newaz D S Sankar Rudy Ramlie Nazmus Shaker Nafi School of Computing and Informatics Universiti Teknologi Brunei (UTB) Gadong Brunei Darussalam KAIST Institute for Information Technology Convergence Daejeon South Korea School of Applied Sciences and Mathematics Universiti Teknologi Brunei (UTB) Gadong Brunei Darussalam Boeing Defence Australia Adelaide SA Australia

Fog computing is paving the way for catering latency stringent applications (e.g., augmented reality and virtual reality) at the customer premises. Fog computing nodes are located between the end user devices and remote cloud. They are lightweight and small-scale storage and processing system deployed closer to the data source, allowing faster processing as well as providing privacy and security of the data. We have been witnessing a growing number of solutions that integrate Passive Optical Network (PON) with the fog computing. Along this research line, there are some solutions considering fog computing nodes are co-located with the Optical Line Terminal (OLT), the central intelligence of a PON system, and Optical Network Unit (ONU), a customer premises equipment. There are some solutions, on the other hand, that propose to embed computing and storage functionality within PON equipment itself. In this paper, we particularly focus on the later approach. Here, we propose an energy conserving solution for a PON system with fog computing enabled ONUs, i.e., the ONUs are equipped with additional computing and Storage Units (CSUs). Our solution aims at minimizing energy consumption of a PON system by keeping only the required number of CSUs active and allowing the ONUs to move into sleep mode (whenever possible) by taking into account task arrival rate and task completion deadline.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Data-Driven Feature Analysis for Student Performance Prediction

Data-Driven Feature Analysis for Student Performance Predict...

引用

Innovative Engineering Sciences and Technological Research (ICIESTR), International Conference on

作者： Maha Said Al-Anqoudi Mireilla Bikanga Ada Stephen McQuistin College of Computing and Information Sciences University of Technology and Applied Science Nizwa Oman School of Computer Science University of Glasgow Glasgow UK School of Computer Science University of St Andrews St Andrews UK

ISBN: (数字)9798350348637

ISBN: (纸本)9798350348644

We have witnessed the widespread adoption of online teaching and learning platforms in recent years. Teachers employ a variety of learning activities and techniques to follow their students’ learning progress, including summative assessments. However, concerns have been raised about the trustworthiness of summative assessments, such as quizzes and tests. At the same time, online teachers have found other learning activities that are not typically included in predictive machine learning models of online platforms are effective in helping them identify student learning status. This study uses two feature selection techniques, filter and embedded, along with a teacher self-reported questionnaire to evaluate which learning activities influence students’ performance in summative assessments like quizzes. Our findings reveal that features from non-summative learning activities can effectively predict students’ performance in an online environment. Teachers agree on the importance of the identified learning activities’ features in helping them determine their students’ learning status. The study provides practical recommendations for educators, course designers, and policy-makers to optimize online assessment strategies.

关键词： Discussion forums Distance learning Education Machine learning Predictive models Feature extraction

来源：评论

学校读者我要写书评

暂无评论

Local meshless methods for elliptic PDEs with multipoint boundary conditions: investigating efficiency and accuracy of various RBFs

引用

The European Physical Journal Special Topics 2024年 1-18页

作者： Ahmad, Masood Khan, Muhammad Nawaz Ahmad, Imtiaz Department of Basic Sciences University of Engineering and Technology Peshawar Peshawar Pakistan Mathematics in Applied Sciences and Engineering Research Group Scientific Research Center Al-Ayen University Nasiriyah Iraq Institute of Informatics and Computing in Energy (IICE) Universiti Tenaga Nasional Kajang Malaysia

The present study addresses the numerical solution of two-dimensional steady-state heat conduction problems with nonlocal multi-point boundary conditions across three distinct domains: a unit rectangle with a quarter-circle cutout of radius 0.5, an irregular domain, and a Cassini curve. Dirichlet boundary conditions are imposed on specific segments, while nonlocal boundary conditions are applied to the remaining portions. The Kansa method is employed to solve the steady-state heat conduction equation, utilizing three types of radial basis functions (RBFs) to explore the influence of the shape parameter on accuracy and matrix conditioning. These include the inverse multiquadric RBF, a modified inverse multiquadric RBF proposed here for the first time, and a hybrid RBF [1]. As a meshless method, the Kansa approach eliminates the need for mesh generation or node connectivity within local subdomains. To evaluate accuracy and performance, the $$L_{\infty }$$ error norm is employed. The results demonstrate the effectiveness of the proposed techniques in solving the 2D steady-state heat conduction problem. A comparative analysis is conducted to assess the accuracy and computational efficiency of the methods.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Advancing Post-growth HCI

Advancing Post-growth HCI

引用

2025 CHI Conference on Human Factors in computing Systems, CHI EA 2025

作者： Sharma, Vishal Lin, Hongjin Wani, Asra Sakeen Katzman, Jared Lee Tuli, Anupriya Karusala, Naveena Bardzell, Shaowen Becker, Christoph Tomitsch, Martin Kumar, Neha School of Interactive Computing Georgia Institute of Technology AtlantaGA United States School of Engineering and Applied Sciences Harvard University AllstonMA United States IIIT Delhi Delhi India School of Information University of Michigan Ann ArborMI United States KTH Royal Institute of Technology Stockholm Sweden Faculty of Information University of Toronto TorontoON Canada Transdisciplinary School University of Technology Sydney SydneyNSW Australia Georgia Tech AtlantaGA United States

ISBN: (纸本)9798400713958

The field of Human-Computer Interaction (HCI) both shapes and is shaped by the forces of economic growth. Extending the calls to move beyond the growth imperative in HCI, this workshop aims to bring HCI researchers, designers, and practitioners together in a critical dialogue on examining the often-hidden ways growth patterns manifest in HCI. We ask how the HCI community can use its research, design, and praxis to build a more emancipatory future. The workshop aims to nurture the Post-growth HCI Collective of scholars and practitioners to work together, in solidarity, to reflect and resist the commodification and drive for infinite capital accumulation in/through digital technologies for the betterment of HCI and the broader computing community. Through collective action, we aim to operationalize post-growth principles in HCI, contributing to more tangible pathways toward socio-ecologically just and sustainable technology-mediated futures for the planet and its residents. © 2025 Copyright held by the owner/author(s).

关键词： Green development

来源：评论

学校读者我要写书评

暂无评论

TrIM: Transformed Iterative Mondrian Forests for Gradient-based Dimension Reduction and High-Dimensional Regression

arXiv

引用

arXiv 2024年

作者： Baptista, Ricardo O’Reilly, Eliza Xie, Yangxinyu Computing and Mathematical Sciences Department California Institute of Technology United States Department of Applied Mathematics and Statistics Johns Hopkins University United States Department of Statistics and Data Science University of Pennsylvania United States

We propose a computationally efficient algorithm for gradient-based linear dimension reduction and high-dimensional regression. The algorithm initially computes a Mondrian forest and uses this estimator to identify a relevant feature subspace of the inputs from an estimate of the expected gradient outer product (EGOP) of the regression function. In addition, we introduce an iterative approach known as Transformed Iterative Mondrian (TrIM) forest to improve the Mondrian forest estimator by using the EGOP estimate to update the set of features and weights used by the Mondrian partitioning mechanism. We obtain consistency guarantees and convergence rates for the estimation of the EGOP matrix and the random forest estimator obtained from one iteration of the TrIM algorithm. Lastly, we demonstrate the effectiveness of our proposed algorithm for learning the relevant feature subspace across a variety of settings with both simulated and real data. Copyright © 2024, The Authors. All rights reserved.

关键词： Iterative methods

来源：评论

学校读者我要写书评

暂无评论

Classification of Attention Performance Post-Longitudinal tDCS via Functional Connectivity and Machine Learning Methods

Classification of Attention Performance Post-Longitudinal tD...

引用

International Conference for Convergence of technology (I2CT)

作者： Akash K Rao Vishnu K Menon Arnav Bhavsar Shubhajit Roy Chowdhury Ramsingh Negi Varun Dutt Applied Cognitive Science Laboratory Indian Institute of Technology Mandi Mandi India School of Computing and Electrical Engineering Indian Institute of Technology Mandi Mandi India Cognitive Control and Machine Learning Group Institute of Nuclear Medicine and Allied Sciences Delhi India

ISBN: (数字)9798350394474

ISBN: (纸本)9798350394481

Attention is the brain's mechanism for selectively processing specific stimuli while filtering out irrelevant information. Characterizing changes in attention following long-term interventions (such as transcranial direct current stimulation (tDCS)) has seldom been emphasized in the literature. To classify attention performance post-tDCS, this study uses functional connectivity and machine learning algorithms. Fifty individuals were split into experimental and control conditions. On Day 1, EEG data was obtained as subjects executed an attention task. From Day 2 through Day 8, the experimental group was administered 1mA tDCS, while the control group received sham tDCS. On Day 10, subjects repeated the task mentioned on Day 1. Functional connectivity metrics were used to classify attention performance using various machine learning methods. Results revealed that combining the Adaboost model and recursive feature elimination yielded a classification accuracy of 91.84%. We discuss the implications of our results in developing neurofeedback frameworks to assess attention.

关键词： Measurement Machine learning algorithms Filtering Machine learning Brain modeling Electroencephalography Task analysis

来源：评论

学校读者我要写书评

暂无评论

Classification of attention performance post-longitudinal tDCS via functional connectivity and machine learning methods

arXiv

引用

arXiv 2024年

作者： Rao, Akash K. Menon, Vishnu K. Bhavsar, Arnav Chowdhury, Shubhajit Roy Negi, Ramsingh Dutt, Varun Applied Cognitive Science Laboratory Indian Institute of Technology Mandi Mandi India School of Computing and Electrical Engineering Indian Institute of Technology Mandi Mandi India Cognitive Control and Machine learning group Institute of Nuclear Medicine and Allied Sciences Delhi India

关键词： Electroencephalography

来源：评论

学校读者我要写书评

暂无评论

Quantum entanglement of microwave and optical photonic qubits with a chip-scale transducer

Quantum entanglement of microwave and optical photonic qubit...

引用

CLEO: Fundamental Science, CLEO:FS 2024 - Part of Conference on Lasers and Electro-Optics, CLEO 2024

作者： Lake, David P. Meesala, Srujan Wood, Steven Chiappina, Piero Zhong, Changchun Beyer, Andrew D. Shaw, Matthew D. Jiang, Liang Painter, Oskar Kavli Nanoscience Institute and Thomas J. Watson Sr. Laboratory of Applied Physics California Institute of Technology PasadenaCA91125 United States Institute for Quantum Information and Matter California Institute of Technology PasadenaCA91125 United States Pritzker School of Molecular Engineering The University of Chicago ChicagoIL60637 United States Jet Propulsion Laboratory California Institute of Technology 4800 Oak Grove Dr PasadenaCA91109 United States Center for Quantum Computing Amazon Web Services PasadenaCA91125 United States

We generate entangled microwave and optical photonic qubits with a chip-scale piezo-optomechanical transducer. We envision such an entangled pair source as an important building block for optical networking of superco... 详细信息

ISBN: (纸本)9781957171395

关键词： Quantum entanglement

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：