检索结果-内蒙古大学图书馆

arXiv 2020年

作者： Park, Jaehyuk Frank, Morgan R. Sun, Lijun Youn, Hyejin Kellogg School of Management Northwestern University EvanstonIL United States Northwestern Institute on Complex Systems EvanstonIL United States Luddy School of Informatics Computing and Engineering Indiana University BloomingtonIN United States Department of Informatics and Networked Systems School of Computing and Information University of Pittsburgh PittsburgePA United States Connection Science Institute for Data Systems and Society Massachusetts Institute of Technology CambridgeMA United States Digital Economy Lab Institute for Human-Centered AI Stanford University StanfordCA United States Department of Civil Engineering McGill University MontrealQC Canada Department of Civil and Environmental Engineering Northwestern University EvanstonIL60208 United States

Categorization is an essential component for us to understand the world for ourselves and to communicate it collectively. It is therefore important to recognize that classification systems is not necessarily static, especially for economic systems, and even more so in urban areas where most innovation takes place and is implemented. Out-of-date classification systems would potentially limit further understanding of the current economy because things constantly change. Here, we develop an occupation-based classification system for the US labor economy, called industrial topics, that satisfy adaptability and representability. By leveraging the distributions of occupations across the US urban areas, we identify industrial topics — clusters of occupations based on their co-existence pattern. Industrial topics indicate the mechanisms under the systematic allocation of different occupations. Considering the densely connected occupations as an industrial topic, our approach characterizes regional economies by their topical composition. Unlike the existing survey-based top-down approach, our method provides timely information about the underlying structure of the regional economy, which is critical for policymakers and business leaders especially in our fast-changing economy. Copyright © 2020, The Authors. All rights reserved.

关键词： Employment

来源：评论

学校读者我要写书评

暂无评论

data-efficient motor imagery decoding in real-time for the cybathlon brain-computer interface race 6

Data-efficient motor imagery decoding in real-time for the c...

引用

6th International Congress on Neurotechnology, Electronics and Informatics, NEUROTECHNIX 2018

作者： Ponferrada, Eduardo G. Sylaidi, Anastasia Aldo Faisal, A. Department of Bioengineering Imperial College London London United Kingdom Department of Computing Imperial College London London United Kingdom Data Science Institute London United Kingdom

ISBN: (纸本)9789897583261

Neuromotor diseases such as Amyotrophic Lateral Sclerosis or Multiple Sclerosis affect millions of people throughout the globe by obstructing body movement and thereby any instrumental interaction with the world. Brain Computer Interfaces (BCIs) hold the premise of re-routing signals around the damaged parts of the nervous system to restore control. However, the field still faces open challenges in training and practical implementation for real-time usage which hampers its impact on patients. The Cybathlon Brain-Computer Interface Race promotes the development of practical BCIs to facilitate clinical adoption. In this work we present a competitive and data-efficient BCI system to control the Cybathlon video game using motor imageries. The platform achieves substantial performance while requiring a relatively small amount of training data, thereby accelerating the training phase. We employ a static band-pass filter and Common Spatial Patterns learnt using supervised machine learning techniques to enable the discrimination between different motor imageries. Log-variance features are extracted from the spatio-temporally filtered EEG signals to fit a Logistic Regression classifier, obtaining satisfying levels of decoding accuracy. The systems performance is evaluated online, on the first version of the Cybathlon Brain Runners game, controlling 3 commands with up to 60.03% accuracy using a two-step hierarchical classifier. Copyright © 2018 by SCITEPRESS - science and Technology Publications, Lda. All rights reserved

关键词： Brain computer interface

来源：评论

学校读者我要写书评

暂无评论

Deep learning for image and point cloud fusion in autonomous driving: A review

arXiv

引用

arXiv 2020年

作者： Cui, Yaodong Chen, Ren Chu, Wenbo Chen, Long Tian, Daxin Li, Ying Cao, Dongpu The Waterloo CogDrive Lab Department of Mechanical and Mechatronics Engineering University of Waterloo 200 University Ave West WaterlooONN2L3G1 Canada Research Institute Co. Ltd. Beijing100176 China School of Data and Computer Science Sun Yat-sen University Zhuhai519082 China Waytous Inc. Qingdao266109 China Beijing Advanced Innovation Center for Big Data and Brain Computing Beijing Key Laboratory for Cooperative Vehicle Infrastructure Systems and Safety Control School of Transportation Science and Engineering Beihang University Beijing100191 China

Autonomous vehicles were experiencing rapid development in the past few years. However, achieving full autonomy is not a trivial task, due to the nature of the complex and dynamic driving environment. Therefore, autonomous vehicles are equipped with a suite of different sensors to ensure robust, accurate environmental perception. In particular, the camera-LiDAR fusion is becoming an emerging research theme. However, so far there has been no critical review that focuses on deep-learning-based camera-LiDAR fusion methods. To bridge this gap and motivate future research, this paper devotes to review recent deep-learning-based data fusion approaches that leverage both image and point cloud. This review gives a brief overview of deep learning on image and point cloud data processing. Followed by in-depth reviews of camera-LiDAR fusion methods in depth completion, object detection, semantic segmentation, tracking and online cross-sensor calibration, which are organized based on their respective fusion levels. Furthermore, we compare these methods on publicly available datasets. Finally, we identified gaps and over-looked challenges between current academic researches and real-world applications. Based on these observations, we provide our insights and point out promising research directions. Copyright © 2020, The Authors. All rights reserved.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Preface

Communications in Computer and Information Science

引用

Communications in Computer and Information science 2025年 2194 CCIS卷 v-vi页

作者： Singh, Mayank Tyagi, Vipin Gupta, P.K. Flusser, Jan Ören, Tuncer Cherif, Amar Ramdane Tomar, Ravi G.L. Bajaj Institute of Technology and Management Uttar Pradesh Greater Noida India Academics and Research Department Jaypee University of Engineering and Technology Madhya Pradesh Raghogarh India School of Computing Department of Data Science and Engineering Mohan Babu University Andhra Pradesh Tirupati India Institute of Information Theory and Automation Prague Czech Republic School of Electrical Engineering and Computer Science OttawaON Canada University of Paris-Saclay Paris France Persistent Systems Uttarakhand Dehradun India

来源：评论

学校读者我要写书评

暂无评论

Cosmology with Galaxy Cluster Properties using Machine Learning

arXiv

引用

arXiv 2023年

作者： Qiu, Lanlan Napolitano, Nicola R. Borgani, Stefano Zhong, Fucheng Li, Xiaodong Radovich, Mario Lin, Weipeng Dolag, Klaus Tortora, Crescenzo Wang, Yang Remus, Rhea-Silvia Wu, Sirui Longo, Giuseppe School of Physics and Astronomy Sun Yat-sen University Zhuhai Campus 2 Daxue Road Tangjia Zhuhai519082 China CSST Science Center for Guangdong-Hong Kong-Macau Great Bay Area Zhuhai519082 China Department of Physics E. Pancini University Federico II Via Cinthia 6 Naples80126-I Italy Astronomy Unit Department of Physics University of Trieste via Tiepolo 11 TriesteI-34131 Italy INAF-Osservatorio Astronomico di Trieste via G. B. Tiepolo 11 TriesteI-34143 Italy IFPU Institute for Fundamental Physics of the Universe Via Beirut 2 Trieste34014 Italy INFN Instituto Nazionale di Fisica Nucleare Via Valerio 2 TriesteI-34127 Italy ICSC - Italian Research Center on High Performance Computing Big Data and Quantum Computing INAF - Osservatorio Astronomico di Padova via dell’Osservatorio 5 Padova35122 Italy Universitäts-Sternwarte Fakultät für Physik Ludwig-Maximilians-Universität München Scheinerstr.1 München81679 Germany Max-Planck-Institut für Astrophysik Karl-Schwarzschild-Straße 1 Garching85741 Germany INAF – Osservatorio Astronomico di Capodimonte Salita Moiariello 16 Napoli80131 Italy Peng Cheng Laboratory No.2 Xingke 1st Street Shenzhen518000 China

Context. Galaxy clusters are the largest gravitating structures in the universe and their mass assembly is sensitive to the underlying cosmology. Their mass function, baryon fraction, and mass distribution have been used to infer cosmological parameters, despite the presence of systematics. However, the complexity of the scaling relations among galaxy cluster properties has never been fully exploited, limiting their potential as a cosmological probe. Aims. We propose the first Machine Learning (ML) method using galaxy cluster properties from hydrodynamical simulations in different cosmologies to predict cosmological parameters combining a series of canonical cluster observables, like gas mass, gas bolometric luminosity, gas temperature, stellar mass, cluster radius, total mass, and velocity dispersion at different redshifts. Methods. The machine learning model is trained on mock "measurements" of these observable quantities from Magneticum multi-cosmology simulations to derive unbiased constraints on a set of cosmological parameters. These include the mass density parameter, Ωm, the power spectrum normalization, σ8, the baryonic density parameter, Ωb, and the reduced Hubble constant, h0. Results. We test the ML model on catalogs of a few hundred clusters taken, in turn, from each simulation and find that the ML model can correctly predict the cosmology they have been picked from. The cumulative accuracy depends on the cosmology, ranging from 21% to 75%. We demonstrate that this is sufficient to derive unbiased constraints on the main cosmological parameters with errors of the order of ∼ 14% for Ωm, ∼ 8% for σ8, ∼ 6% for Ωb, and ∼ 3% for h0. Conclusions. This proof-of-concept analysis, yet based on a limited variety of multi-cosmology simulations, shows that machine learning can efficiently map the correlations in the multi-dimensional space of the observed quantities to the cosmological parameter space and narrow down the probability that a given sample belongs to a

关键词： Cosmology

来源：评论

学校读者我要写书评

暂无评论

Design and Implementation of virtual simulation experiment platform for computer specialized courses

引用

Journal of Physics: Conference Series 2020年第1期1693卷

作者： Guanghui Song Yan Nie Genlang Chen Youcheng Tong School of Computing and Data Engineering Zheda Ningbo Institute of Technology Ningbo 315100 China College of Science and Technology Ningbo University Ningbo 315000 China

For the practical teaching of computer specialized courses in colleges and universities, this study proposes the construction method of virtual simulation experiment platform based on OpenStack, Docker, Seafile and Raspberry Pi, and on this basis, deeply discuss the design and implementation details of the course experiment. The construction and application of virtual simulation experiment platform change the traditional teaching method fundamentally, reduce the cost of laboratory construction and management, which have important reference value to realize the remote experiment teaching. The virtual simulation experiment platform is composed of four parts: virtual machine, application container, cloud storage and client, which can meet the daily teaching needs, cultivate students' interest in learning, and improve their innovation ability.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Hyperparameter Estimation in SVM with GPU Acceleration for Prediction of Protein-Protein Interactions

Hyperparameter Estimation in SVM with GPU Acceleration for P...

引用

IEEE International Conference on Big data

作者： Huaming Chen Lei Wang Yaochu Jin Chi-Hung Chi Fucun Li Huaiyuan Chu Jun Shen School of Computing and Information Technology University of Wollongong Wollongong NSW Australia Department of Computer Science University of Surrey United Kingdom Data61 CSIRO Australia Rensselaer Polytechnic Institute United States

ISBN: (数字)9781728108582

ISBN: (纸本)9781728108599

For classification tasks, such as protein-protein interactions (PPI), support vector machines (SVMs) have been continually utilised as a standard machine learning model. However, most practices in PPIs classifications are limited to common circumstances with small datasets and low feature dimensions, due to the big computation burden of kernel functions and quadratic optimization of SVM. Alternatively, these practical experiences might tend to employ a linear model once the dataset becomes larger, which may have exclusively lost the kernel function's potential. Since there are different defined kernels and various groups of hyperparameter, the time costs in estimating a best set of hyperparameter by traditional grid search are subsequently tremendous for PPI classification. To address this challenge, in this paper, we present a more efficient solution of hyperparameter estimation by gaining acceleration with GPU, which trains SVM efficiently and accurately with kernel functions calculation accelerated on various PPI datasets. The experiments are firstly conducted on PPI classification task, and we have exclusively evaluated the effectiveness on five public classification datasets. Our solution demonstrates a faster and more accurate performance comparing with the state-of-the-art.

关键词： Kernel Support vector machines Graphics processing units Task analysis Estimation Training Acceleration

来源：评论

学校读者我要写书评

暂无评论

HAS-QA: Hierarchical answer spans model for open-domain question answering

arXiv

引用

arXiv 2019年

作者： Pang, Liang Lan, Yanyan Guo, Jiafeng Xu, Jun Su, Lixin Cheng, Xueqi CAS Key Laboratory of Network Data Science and Technology Institute of Computing Technology Chinese Academy of Sciences Beijing China University of Chinese Academy of Sciences Beijing China Department of Statistics University of California Berkeley

This paper is concerned with open-domain question answering (i.e., OpenQA). Recently, some works have viewed this problem as a reading comprehension (RC) task, and directly applied successful RC models to it. However, the performances of such models are not so good as that in the RC task. In our opinion, the perspective of RC ignores three characteristics in OpenQA task: 1) many paragraphs without the answer span are included in the data collection;2) multiple answer spans may exist within one given paragraph;3) the end position of an answer span is dependent with the start position. In this paper, we first propose a new probabilistic formulation of OpenQA, based on a three-level hierarchical structure, i.e., the question level, the paragraph level and the answer span level. Then a Hierarchical Answer Spans Model (HAS-QA) is designed to capture each probability. HAS-QA has the ability to tackle the above three problems, and experiments on public OpenQA datasets show that it significantly outperforms traditional RC baselines and recent OpenQA baselines. Copyright © 2019, The Authors. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

First Detection of the Monkeypox Virus Using Wastewater-Based Surveillance in Miami-Dade County

SSRN

引用

SSRN 2022年

作者： Sharkey, Mark E. Babler, Kristina M. Abelson, Samantha M. Alsuliman, Bader Amirali, Ayaaz Comerford, Samuel Grills, George S. Kumar, Naresh Laine, Jennifer Lamar, Walter E. Mason, Christopher E. Penso, Johnathon Reding, Brian D. Schürer, Stephan C. Shukla, Bhavarth S. Stevenson, Mario Vidovic, Dusica Solo-Gabriele, Helena M. Department of Medicine University of Miami Miller School of Medicine MiamiFL United States Department of Chemical Environmental and Materials Engineering University of Miami Coral GablesFL United States Department of Public Health Sciences University of Miami Miller School of Medicine MiamiFL United States Sylvester Comprehensive Cancer Center University of Miami Miller School of Medicine MiamiFL United States Environmental Health and Safety University of Miami MiamiFL United States Facilities Safety & Compliance University of Miami Miller School of Medicine MiamiFL United States Department of Physiology and Biophysics Weill Cornell Medical College New York CityNY United States Department of Molecular & Cellular Pharmacology University of Miami Miller School of Medicines MiamiFL United States Institute for Data Science & Computing University of Miami Coral GablesFL United States

Measurements of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) RNA from wastewater have been utilized for evaluating the prevalence of COVID-19 illness within communities and other sites, such as institutions of higher learning. A wastewater-based surveillance program was established at the University of Miami to monitor SARS-CoV-2 RNA levels and subsequently exploited to detect both RNA and DNA of additional biological targets present in concentrated wastewater samples. Here we report on the use of wastewater to detect nucleic acids of Monkeypox virus (MPXV) which has recently emerged to become a concern worldwide. Results show that MPXV can be detected in wastewater based on PCR amplification of both MPXV DNA and RNA. Levels of MPXV nucleic acids in wastewater generally reflect the trends of clinical cases in the community and since its initial detection, MPXV levels have remained above PCR detection limits throughout the sampling period. We recommend the expansion of current wastewater-based programs’ methods to detect a broader range of pathogens of concern in wastewater and propose complementary approaches that target both DNA and RNA to increase the sensitivity of detecting disease causing pathogens within communities. © 2022, The Authors. All rights reserved.

关键词： Coronavirus

来源：评论

学校读者我要写书评

暂无评论

494P Evaluation of transcriptome forward computational strategies to improve molecular diagnosis in rare pediatric neuromuscular disease

引用

Neuromuscular Disorders 2024年 43卷 104441.331-104441.331页

作者： Silverstein, S. Ganapathy, K. Donkervoort, S. Moy, J. Uapinyoying, B. Gorokhova, S. Ganesh, V. Weisburd, B. Orbach, R. Foley, A. Mohammadi, P. Adams, D. Bonnemann, C. Neuromuscular and Neurogenetics Disorders of Childhood Section NINDS NIH Bethesda USA Undiagnosed Diseases Program NIH Bethesda USA Rutgers New Jersey Medical School Newark USA Department of Integrative Structural and Computational Biology Scripps Research Institute La Jolla USA Graduate College of Computing and Data Science Boston University Boston USA Center for Genetic Medicine Research Children's Research Institute Children's National Hospital Washington DC USA Aix Marseille Univ INSERM MMG Marseille France Department of Medical Genetics Timone Children's Hospital Marseille France The Broad Institute of MIT and Harvard Boston USA Center for Immunity and Immunotherapies Seattle Children's Research Institute Seattle USA Department of Pediatrics University of Washington School of Medicine Seattle USA Department of Genome Sciences University of Washington Seattle USA

Pediatric neuromuscular diseases are a genetically and clinically heterogenous group of disorders of which 30-60% remain molecularly undiagnosed even after evaluation with exome or genome sequencing. Estimates for added diagnostic benefit of transcriptomic sequencing range between 7-32% across different disease groups. Recently, open-source computational tools designed to systematically survey transcriptomic sequencing data for aberrant events have become available. In this study we assess the sensitivity of DROP (OUTRIDER, FRASER and MAE), MINTIE, LeafcutterMD, and ANEVA-H to detect known outliers in pediatric neuromuscular cohorts in order to identify the optimal set of tools for diagnostic use. Each tool selected utilizes a different statistical strategy and experimental design geared for rare disease. Tools were run on three transcriptome cohorts from the NINDS Neuromuscular and Neurogenetic Diseases of Childhood Section: muscle tissue (n=71), cycloheximide treated fibroblasts (n=79) and untreated fibroblasts (n=37). For splicing tools, MINTIE is the best performing tool (recall rate = 9/21, 42%), followed by FRASER (recall rate = 16/41, 36%), then LeafcutterMD (recall rate = 0/41, 0%). For expression outliers, OUTRIDER detects 6% of true positives (1/15), while the addition of healthy GTEx controls does not improve detection. For monoallelic expression, MAE and ANEVA-H will be compared. Additional evaluations to identify the optimal RNA library preparation methodology for these tools is also in progress. Thus far, OUTRIDER yielded one new candidate diagnosis of TRIP4 related disease. Analysis for new candidate variants is ongoing. Based on these results, and those of ongoing experiments, we will provide recommendations for the use of these tools in diagnostic pipelines for rare pediatric neuromuscular disease.

关键词：

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：