检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Kanatsoulis, Charilaos I. Choi, Evelyn Jegelka, Stephanie Leskovec, Jure Ribeiro, Alejandro Dept. of Computer Science Stanford University United States School of Computation Information and Technology Technical University of Munich Computer Science & Artificial Intelligence Laboratory Massachusetts Institute of Technology United States Dept. of Electrical and Systems Engineering University of Pennsylvania United States

Positional encodings (PEs) are essential for effective graph representation learning because they provide position awareness in inherently position-agnostic transformer architectures and increase the expressive capacity of Graph Neural Networks (GNNs). However, designing powerful and efficient PEs for graphs poses significant challenges due to the absence of canonical node ordering and the scale of the graph. In this work, we identify four key properties that graph PEs should satisfy: stability, expressive power, scalability, and genericness. We find that existing eigenvector-based PE methods often fall short of jointly satisfying these criteria. To address this gap, we introduce PEARL, a novel framework of learnable PEs for graphs. Our primary insight is that message-passing GNNs function as nonlinear mappings of eigenvectors, enabling the design of GNN architectures for generating powerful and efficient PEs. A crucial challenge lies in initializing node attributes in a manner that is both expressive and permutation equivariant. We tackle this by initializing GNNs with random node inputs or standard basis vectors, thereby unlocking the expressive power of message-passing operations, while employing statistical pooling functions to maintain permutation equivariance. Our analysis demonstrates that PEARL approximates equivariant functions of eigenvectors with linear complexity, while rigorously establishing its stability and high expressive power. Experimental evaluations show that PEARL outperforms lightweight versions of eigenvector-based PEs and achieves comparable performance to full eigenvector-based PEs, but with one or two orders of magnitude lower complexity. Our code is available at https://***/ehejin/Pearl-PE. Copyright © 2025, The Authors. All rights reserved.

关键词： Encoding (symbols)

来源：评论

学校读者我要写书评

暂无评论

Novel Distributed Informatics Platform to Support Machine Learning Discovery of Metabolic Biomarkers in Hypoxia Predisposition

Novel Distributed Informatics Platform to Support Machine Le...

引用

16th International Joint Conference on Biomedical Engineering systems and Technologies, BIOSTEC 2023

作者： Stell, Anthony Chauhan, Vedant Amador, Sandra Beuschlein, Felix Favier, Judith Gil, David Greenwood, Philip de Krijger, Ronald Kroiss, Matthias Ortuno, Samanta Patocs, Attila Walch, Axel School of Computing and Information Systems University of Melbourne Melbourne Australia Zurich Switzerland Universite Paris Cite PARCC INSERM Equipe Labellisee par la Ligue contre le Cancer Paris France Department of Computer Science Technology and Computation University of Alicante Alicante Spain Dept. of Pathology University Medical Center Utrecht and Princess Maxima Center for Pediatric Oncology Utrecht Netherlands University Hospital Munich Ludwig-Maximilians-Universitat Munchen Germany ELKH Hereditary Cancer Research Group Department of Laboratory Medicine Semmelweis University and Department of Molecular Genetics National Institute of Oncology Budapest Hungary Research Unit Analytical Pathology Helmholtz Zentmm Munchen Oberschleissheim Germany Department of Computer Science and Systems University of Murcia Murcia Spain Health and Biomedical Research Institute of Alicante Alicante Spain

To realise the scientific and clinical benefits of machine learning (ML) in a multi-centre research collaboration, a common issue is tire need to bring high-volnme data, complex analytical algorithms, and large-scale processing power, all together into one place. This paper describes tire detailed architecture of a novel platform that combines these features, in tire context of a proposed new clinical/bioinformatics project, Hypox-PD. Hypox-PD uses ML methods to identify new metabolic bionrarkers, through the analysis of high-volume data including mass spectrometry and imaging morphology of biobank tissue. The platform features three components: a content delivery network (CDN);a standardised orchestration application;and lrigh-specification processing power and storage. The central innovation of this platform is a distributed application that simultaneously manages the workflow between these components, provides a virtual mapping of the domain data dictionary, and presents the project data/metadata in a FAIR-compliant external interface. This paper presents the detailed design specifications of this platform, as well as initial test results in establishing the benchmark challenge of current direct transfer times without any specialised support. An initial costing of CDN usage is also presented, which indicates that significant performance improvement may be achievable at a reasonable cost to research budgets. © 2023 by SCITEPRESS – Science and Technology Publications, Lda.

关键词： Budget control

来源：评论

学校读者我要写书评

暂无评论

IEEE ICCI*CC Series in Year 20: Latest Advances in Cognitive Computing (Plenary Panel Report-II) 20

IEEE ICCI*CC Series in Year 20: Latest Advances in Cognitive...

引用

20th IEEE International Conference on Cognitive Informatics and Cognitive Computing, ICCI*CC 2021

作者： Kinsner, Witold Zhu, Haibin Baciu, George Luo, Guiming Rubio, Fernando Sui, Jie Huang, Runhe Hiraishi, Hironori Peng, Jun Chen, Liang Univ. of Manitoba FI2CICC Dept of ECE Winnipeg Canada Nipissing University FI2CICC ON Canada Hong Kong Polytechnic University FI2CICC Dept. of Computing Hong Kong Tsinghua University FI2CICC Software Institute Beijing China Complutense Univ. of Madrid FI2CICC Dept. of Computer Systems and Computation Spain Aberdeen University FI2CICC Aberdeen United Kingdom Hosei University FI2CICC Faculty of Computer and Information Sciences Japan Ashikaga University Faculty of Engineering Tochigi326-8558 Japan Chongqing Univ. of Science and Technology FI2CICC School of AI Chongqing China Univ. of Northern British Columbia Dept. of Computer Science BC Canada

ISBN: (纸本)9781665421195

Cognitive Computing (CC) is a contemporary field of fundamental intelligence theories and general AI technologies triggered by the transdisciplinary development in intelligence, computer, brain, knowledge, cognitive, robotic, and cybernetic sciences for engineering implementations. This paper presents a summary of the plenary panel (Part II) on the theoretical foundations of CI/CC and recent breakthroughs in AI engineering reported in the 20th IEEE International ICCI*CC Conference (ICCI*CC'21). The latest advances in CI and CC towards general AI are presented by twenty-Two distinguished panelists. Strategic AI engineering applications in CI, CC, and cognitive systems are elaborated for abstract intelligence, cognitive robots, autonomous systems, intelligent vehicles, and safety-And-mission-critical systems. © 2021 IEEE.

关键词： Cognitive systems

来源：评论

学校读者我要写书评

暂无评论

IEEE ICCI*CC Series in Year 20: Latest Advances in Cognitive Computing (Plenary Panel Report-II)

IEEE ICCI*CC Series in Year 20: Latest Advances in Cognitive...

引用

IEEE International Conference on Cognitive Informatics

作者： Witold Kinsner Haibin Zhu George Baciu Guiming Luo Fernando Rubio Jie Sui Runhe Huang Hironori Hiraishi Jun Peng Liang Chen FI2CICC Dept of ECE Univ. of Manitoba Winnipeg Canada FI2CICC Nipissing University ON Canada FI2CICC Dept. of Computing Hong Kong Polytechnic University Hong Kong FI2CICC Software Institute Tsinghua University Beijing China FI2CICC Dept. of Computer Systems &#x0026 Computation Complutense Univ. of Madrid Spain FI2CICC Aberdeen University Aberdeen UK FI2CICC Faculty of Computer and Information Sciences Hosei University Japan Faculty of Engineering Ashikaga University Tochigi Japan FI2CICC School of AI Chongqing Univ. of Science and Technology Chongqing China Dept. of Computer Science Univ. of Northern British Columbia BC Canada

ISBN: (纸本)9781665421201

Cognitive Computing (CC) is a contemporary field of fundamental intelligence theories and general AI technologies triggered by the transdisciplinary development in intelligence, computer, brain, knowledge, cognitive, robotic, and cybernetic sciences for engineering implementations. This paper presents a summary of the plenary panel (Part II) on the theoretical foundations of CI/CC and recent breakthroughs in AI engineering reported in the 20th IEEE International ICCI*CC Conference (ICCI*CC'21). The latest advances in CI and CC towards general AI are presented by twenty-two distinguished panelists. Strategic AI engineering applications in CI, CC, and cognitive systems are elaborated for abstract intelligence, cognitive robots, autonomous systems, intelligent vehicles, and safety-and-mission-critical systems.

关键词： Knowledge engineering Intelligent vehicles Autonomous systems Conferences Cognitive informatics Cognitive robotics Cognitive systems

来源：评论

学校读者我要写书评

暂无评论

Redundant Node Management in Wireless Sensor Networks with Multiple Sensor Types

Redundant Node Management in Wireless Sensor Networks with M...

引用

2018 National Information Technology Conference, NITC 2018

作者： Maduranga, M.W.P. Saengudomlert, P. Dilum Bandara, H.M.N. Dept. of Computer Science and Engineering University of Moratuwa Katubedda Sri Lanka Bangkok University Center of Research in Optoelectronics Communications and Computation Systems Bangkok University Pathum Thani Thailand

ISBN: (纸本)9781538691366

One of the major challenges in Wireless Sensor Networks (WSNs) is saving energy to prolonging sensor motes' lifetime, as battery capacities are limited and often difficult to replace or re-charge. In dense sensor a mote deployment, which is often the case, energy consumption could be reduced by activating only a subset of the sensors at a time based on the spatial and temporal correlations among the sensors. However, identifying such subsets is nontrivial as the associated set-cover problem is NP-complete. Therefore, we propose a heuristic-based algorithm called the Correlated Set Calculation Algorithm (CSCA) to determine the subset of motes to be activated at a given time. CSCA considers both the spatial correlation among sensors and multiple sensor types in determining the subset of sensor motes to activate. Therefore, CSCA could enhance the coverage of the area of interest while minimizing the overall energy expenditure and maximizing the lifetime of the WSN. Experimental and simulation results show that a significant number of motes can be turned off based on the schedule calculated by CSCA. Thus, the proposed algorithm could save significant energy and prolong the lifetime of sensors extensively. © 2018 IEEE.

关键词： Energy utilization

来源：评论

学校读者我要写书评

暂无评论

Infrastructure based on template engines for automatic generation of source code for self-Adaptive software domain 29

Infrastructure based on template engines for automatic gener...

引用

29th International Conference on Software Engineering and Knowledge Engineering, SEKE 2017

作者： Benato, Gabriele Salgado Affonso, Frank José Nakagawa, Elisa Yumi Dept. of Statistics Applied Mathematics and Computation São Paulo State University - UNESP PO Box 178 Rio Claro SP13506-900 Brazil Dept. of Computer Systems University of São Paulo - USP PO Box 668 São Carlos SP13560-970 Brazil

ISBN: (纸本)1891706411

Self-Adaptive Software (SaS), a special class of software system, constantly deals with some type of changes (i.e., structural and/or behavioral) to meet the user's new needs or autonomously react to modifications in its execution environment. Software adaptation, when manually performed, becomes an onerous and error-prone activity. Automated approaches have been adopted as a feasible alternative to overcome such adversities because they reduce the human involvement to a minimum. Based on this scenario, a Reference Architecture for SaS (RA4SaS) [1], [2] was designed in previous work. This architecture enables the adaptation of SaS at runtime by means of a controlled adaptation approach. In this sense, an automated process and a complete infrastructure to allow the phases of such process were also developed. This paper presents the design and implementation of a new module for automatic generation of source code for our architecture based on template engines [3]. In short, the main benefits of this module can be summarized in three items: (i) higher design flexibility, maintenance and evolution of SaS;(ii) introduction of new capabilities for automatic generation of source code;and (iii) performance improvement. To present the operation of this new module, a case study was conducted. As result, we have observed that our architecture has good perspective to efficiently contribute to the SaS area.

关键词： Automation

来源：评论

学校读者我要写书评

暂无评论

Bernoulli random forests: Closing the gap between theoretical consistency and empirical soundness 25

Bernoulli random forests: Closing the gap between theoretica...

引用

25th International Joint Conference on Artificial Intelligence, IJCAI 2016

作者： Yisen, Wang Qingtao, Tang Xia, Shu-Tao Wu, Jia Zhu, Xingquan Dept. of Computer Science and Technology Tsinghua University China Graduate School at Shenzhen Tsinghua University China Quantum Computation and Intelligent Systems Centre University of Technology Sydney Australia Dept. of Computer and Electrical Engineering and Computer Science Florida Atlantic University United States

Random forests are one type of the most effective ensemble learning methods. In spite of their sound empirical performance, the study on their theoretical properties has been left far behind. Recently, several random forests variants with nice theoretical basis have been proposed, but they all suffer from poor empirical performance. In this paper, we propose a Bernoulli random forests model (BRF), which intends to close the gap between the theoretical consistency and the empirical soundness of random forests classification. Compared to Breiman's original random forests, BRF makes two simplifications in tree construction by using two independent Bernoulli distributions. The first Bernoulli distribution is used to control the selection of candidate attributes for each node of the tree, and the second one controls the splitting point used by each node. As a result, BRF enjoys proved theoretical consistency, so its accuracy will converge to optimum (i.e., the Bayes risk) as the training data grow infinitely large. Empirically, BRF demonstrates the best performance among all theoretical random forests, and is very comparable to Breiman's original random forests (which do not have the proved consistency yet). The theoretical and experimental studies advance the research one step further towards closing the gap between the theory and the practical performance of random forests classification.

关键词： Forestry

来源：评论

学校读者我要写书评

暂无评论

DmS-Modeler: A tool for modeling decision-making systems for self-adaptive software domain 28

DmS-Modeler: A tool for modeling decision-making systems for...

引用

28th International Conference on Software Engineering and Knowledge Engineering, SEKE 2016

作者： Affonso, Frank José Leite, Gustavo Nakagawa, Elisa Yumi Dept. of Statistics Applied Mathematics and Computation Univ Estadual Paulista - UNESP Rio ClaroSP Brazil Dept. of Computer Systems University of São Paulo - USP São Carlos SP Brazil

ISBN: (纸本)189170639X

The ability to modify its own structure and/or behavior at runtime is a native feature in the development of Self-adaptive Software (SaS). In previous work, a Reference Architecture for SaS (RA4SaS), an automated process for adaptation, and a framework for decision-making were developed to assist the development of SaS. Although such initiatives have collaborated with evolution of SaS, the design of the Decisionmaking systems (DmS), element of first class for SaS, is manually conducted. Therefore, this paper presents a tool called DmSModeler, which aims to assist the development of DmS for SaS, providing facilities for modeling, calibration of such system, and automatic generation of infrastructure (i.e., source code and databases). Aiming to present the applicability of our tool, a case study was conducted and the results enable us to have good perspectives of contribution to the SaS area and other domains of software systems.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

A framework based on learning techniques for decision-making in self-adaptive software 27

A framework based on learning techniques for decision-making...

引用

27th International Conference on Software Engineering and Knowledge Engineering, SEKE 2015

作者： Affonso, Frank José Leite, Gustavo Oliveira, Rafael A.P. Nakagawa, Elisa Yumi Dept. of Statistics Applied Mathematics and Computation Univ Estadual Paulista-UNESP Rio-Claro SP Brazil Dept. of Computer Systems University of São Paulo-USP São-Carlos SP Brazil

ISBN: (纸本)1891706373

The development of Self-adaptive Software (SaS) presents specific innovative features compared to traditional ones since this type of software constantly deals with structural and/or behavioral changes at runtime. Capabilities of human administration are showing a decrease in relative effectiveness, since some tasks have been difficult to manage introducing potential problems, such as change management and simple human error. Self-healing systems, a system class of SaS, have emerged as a feasible solution in contrast to management complexity, since such system often combines machine learning techniques with control loops to reduce the number of situations requiring human intervention. This paper presents a framework based on learning techniques and the control loop (MAPE-K) to support the decision-making activity for SaS. In addition, it is noteworthy that this framework is part of a wider project developed by the authors of this paper in previous work (i.e., reference architecture for SaS [1]). Aiming to present the viability of our framework, we have conducted a case study using a flight plan module for Unmanned Aerial Vehicles. The results have shown an environment accuracy of about 80%, enabling us to project good perspectives of contribution to the SaS area and other domains of software systems, and enabling knowledge sharing and technology transfer from academia to industry. Copyright © 2015 by KSI Research Inc. and Knowledge systems Institute Graduate School.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

Challenging local realism with human choices

arXiv

引用

arXiv 2018年

作者： Abellán, C. Acín, A. Alarcón, A. Alibart, O. Andersen, C.K. Andreoli, F. Beckert, A. Beduini, F.A. Bendersky, A. Bentivegna, M. Bierhorst, P. Burchardt, D. Cabello, A. Cariñe, J. Carrasco, S. Carvacho, G. Cavalcanti, D. Chaves, R. Cortés-Vega, J. Cuevas, A. Delgado, A. de Riedmatten, H. Eichler, C. Farrera, P. Fuenzalida, J. García-Matos, M. Garthoff, R. Gasparinetti, S. Gerrits, T. GhafariJouneghani, F. Glancy, S. Gómez, E.S. González, P. Guan, J.-Y. Handsteiner, J. Heinsoo, J. Heinze, G. Hirschmann, A. Jiménez, O. Kaiser, F. Knill, E. Knoll, L.T. Krinner, S. Kurpiers, P. Larotonda, M.A. Larsson, J.-Å. Lenhard, A. Li, H. Li, M.-H. Lima, G. Liu, B. Liu, Y. López Grande, I.H. Lunghi, T. Ma, X. Magaña-Loaiza, O.S. Magnard, P. Magnoni, A. Martí-Prieto, M. Martínez, D. Mataloni, P. Mattar, A. Mazzera, M. Mirin, R.P. Mitchell, M.W. Nam, S. Oppliger, M. Pan, J.-W. Patel, R.B. Pryde, G.J. Rauch, D. Redeker, K. Rieländer, D. Ringbauer, M. Roberson, T. Rosenfeld, W. Salathé, Y. Santodonato, L. Sauder, G. Scheidl, T. Schmiegelow, C.T. Sciarrino, F. Seri, A. Shalm, L.K. Shi, S.-C. Slussarenko, S. Stevens, M.J. Tanzilli, S. Toledo, F. Tura, J. Ursin, R. Vergyris, P. Verma, V.B. Walter, T. Wallraff, A. Wang, Z. Weinfurter, H. Weston, M.M. White, A.G. Wu, C. Xavier, G.B. You, L. Yuan, X. Zeilinger, A. Zhang, Q. Zhang, W. Zhong, J. 08860 Spain ICREA Institució Catalana de Recerca i Estudis Avançats Barcelona08010 Spain Center for Optics and Photonics MSI-Nucleus for Advanced Optics Universidad de Concepción Casilla 4016 Concepción Chile Departamento de Ingeniería Eléctrica Universidad de Concepción 160-C Concepción Chile Parc Valrose Nice Cedex 206108 France Quantum Device Lab Department of Physics ETH Zurich ZurichCH-8093 Switzerland Dipartimento di Fisica Sapienza Università di Roma RomaI-00185 Italy Departamento de Computación FCEyN UBA and ICC CONICET Ciudad Universitaria Pabellón 1 Buenos Aires1428 Argentina National Institute of Standards and Technology 325 Broadway BoulderCO80305 United States Ludwig-Maximilians-Universität München80799 Germany Departamento de Física Aplicada II Universidad de Sevilla Sevilla41012 Spain Departamento de Física Universidad de Concepción 160-C Concepción Chile International Institute of Physics Federal University of Rio Grande do Norte Natal59070-405 Brazil Austrian Academy of Sciences Boltzmanngasse 3 Vienna1090 Austria Centre for Quantum Computation and Communication Technology Griffith University BrisbaneQLD4111 Australia Centre for Quantum Dynamics Griffith University BrisbaneQLD4111 Australia Shanghai Branch National Laboratory for Physical Sciences at Microscale Dept. of Modern Physics University of Science and Technology of China Shanghai230026 China DEILAP CITEDEF & CONICET J.B. de La Salle 4397 1603 Villa Martelli Buenos Aires Argentina Departamento de Física FCEyN UBA and IFIBA Ciudad Universitaria Conicet Pabellón 1 Buenos Aires1428 Argentina Department of Electrical Engineering Linköping University Linköping581 83 Sweden State Key Laboratory of Functional Materials for Informatics Shanghai Institute of Microsystem and Information Technology Chinese Academy of Sciences Shanghai200050 China CAS Center for Excellence in Superconducting Electronics Shanghai200050 China School of Computer NUDT

A Bell test is a randomized trial that compares experimental observations against the philosophical worldview of local realism1, in which the properties of the physical world are independent of our observation of them and no signal travels faster than light. A Bell test requires spatially distributed entanglement, fast and high-efficiency detection and unpredictable measurement settings2,3. Although technology can satisfy the first two of these requirements4–7, the use of physical devices to choose settings in a Bell test involves making assumptions about the physics that one aims to test. Bell himself noted this weakness in using physical setting choices and argued that human ‘free will’ could be used rigorously to ensure unpredictability in Bell tests8. Here we report a set of local-realism tests using human choices, which avoids assumptions about predictability in physics. We recruited about 100,000 human participants to play an online video game that incentivizes fast, sustained input of unpredictable selections and illustrates Bell-test methodology9. The participants generated 97,347,490 binary choices, which were directed via a scalable web platform to 12 laboratories on five continents, where 13 experiments tested local realism using photons5,6, single atoms7, atomic ensembles10, and superconducting devices11. Over a 12-hour period on 30 November 2016, participants worldwide provided a sustained data flow of over 1,000 bits per second to the experiments, which used different human-generated data to choose each measurement setting. The observed correlations strongly contradict local realism and other realistic positions in bipartite and tripartite12 scenarios. Project outcomes include closing the ‘freedom-of-choice loophole’ (the possibility that the setting choices are influenced by ‘hidden variables’ to correlate with the particle properties13), the utilization of video-game methods14 for rapid collection of human generated randomness, and the use of network

关键词： Testing

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：