检索结果-内蒙古大学图书馆

Inferring Human Intent and Predicting Human Action in Human–Robot Collaboration

Control, Robotics, and Autonomous Systems 1000年第1期7卷 73-95页

作者： Guy Hoffman Tapomayukh Bhattacharjee Stefanos Nikolaidis 1Sibley School of Mechanical and Aerospace Engineering Cornell University Ithaca New York USA email: hoffman@cornell.edu 2Department of Computer Science Cornell University Ithaca New York USA 3Thomas Lord Department of Computer Science University of Southern California Los Angeles California USA

Researchers in human–robot collaboration have extensively studied methods for inferring human intentions and predicting their actions, as this is an important precursor for robots to provide useful assistance. We review contemporary methods for intention inference and human activity prediction. Our survey finds that intentions and goals are often inferred via Bayesian posterior estimation and Markov decision processes that model internal human states as unobserved variables or represent both agents in a shared probabilistic framework. An alternative approach is to use neural networks and other supervised learning approaches to directly map observable outcomes to intentions and to make predictions about future human activity based on past observations. That said, due to the complexity of human intentions, existing work usually reasons about limited domains, makes unrealistic simplifications about intentions, and is mostly constrained to short-term predictions. This state of the art provides opportunity for future research that could include more nuanced models of intents, reason over longer horizons, and account for the human tendency to adapt.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Decision Analysis for Management of Natural Hazards

引用

Environment and Resources 1000年第1期41卷 489-516页

作者： Michael Simpson Rachel James Jim W. Hall Edoardo Borgomeo Matthew C. Ives Susana Almeida Ashley Kingsborough Theo Economou David Stephenson Thorsten Wagener 1Environmental Change Institute University of Oxford Oxford OX1 3QY United Kingdom email: michael.simpson@ouce.ox.ac.uk rachel.james@ouce.ox.ac.uk jim.hall@eci.ox.ac.uk edoardo.borgomeo@ouce.ox.ac.uk matthew.ives@ouce.ox.ac.uk ashley.kingsborough@ouce.ox.ac.uk 2Department of Civil Engineering University of Bristol Bristol BS8 1TR United Kingdom email: susana.almeida@bristol.ac.uk thorsten.wagener@bristol.ac.uk 3Department of Mathematics and Computer Science University of Exeter Exeter EX4 4QF United Kingdom email: t.economou@exeter.ac.uk d.b.stephenson@exeter.ac.uk 4Cabot Institute Royal Fort House University of Bristol Bristol BS8 1UJ United Kingdom

Losses from natural hazards, including geophysical and hydrometeorological hazards, have been increasing worldwide. This review focuses on the process by which scientific evidence about natural hazards is applied to support decision making. Decision analysis typically involves estimating the probability of extreme events; assessing the potential impacts of those events from a variety of perspectives; and evaluating options to plan for, mitigate, or react to events. We consider issues that affect decisions made across a range of natural hazards, summarize decision methodologies, and provide examples of applications of decision analysis to the management of natural hazards. We conclude that there is potential for further exchange of ideas and experience between natural hazard research communities on decision analysis approaches. Broader application of decision methodologies to natural hazard management and evaluation of existing decision approaches can potentially lead to more efficient allocation of scarce resources and more efficient risk management.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Machine Learning for Sustainable Energy Systems

引用

Environment and Resources 1000年第1期46卷 719-747页

作者： Priya L. Donti J. Zico Kolter 2Department of Engineering and Public Policy Carnegie Mellon University Pittsburgh Pennsylvania 15213 USA 1Computer Science Department Carnegie Mellon University Pittsburgh Pennsylvania 15213 USA email: pdonti@cs.cmu.edu 3Bosch Center for Artificial Intelligence Pittsburgh Pennsylvania 15222 USA

In recent years, machine learning has proven to be a powerful tool for deriving insights from data. In this review, we describe ways in which machine learning has been leveraged to facilitate the development and operation of sustainable energy systems. We first provide a taxonomy of machine learning paradigms and techniques, along with a discussion of their strengths and limitations. We then provide an overview of existing research using machine learning for sustainable energy production, delivery, and storage. Finally, we identify gaps in this literature, propose future research directions, and discuss important considerations for deployment.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Toward Realizing the Promise of AI in Precision Health Across the Spectrum of Care

引用

Genomics and Human Genetics 1000年第1期25卷 141-159页

作者： Jenna Wiens Kayte Spector-Bagdady Bhramar Mukherjee 1Division of Computer Science and Engineering College of Engineering University of Michigan Ann Arbor Michigan USA email: wiensj@umich.edu 2Department of Obstetrics and Gynecology and Center for Bioethics and Social Sciences in Medicine University of Michigan Medical School Ann Arbor Michigan USA 3Department of Biostatistics School of Public Health University of Michigan Ann Arbor Michigan USA

Significant progress has been made in augmenting clinical decision-making using artificial intelligence (AI) in the context of secondary and tertiary care at large academic medical centers. For such innovations to have an impact across the spectrum of care, additional challenges must be addressed, including inconsistent use of preventative care and gaps in chronic care management. The integration of additional data, including genomics and data from wearables, could prove critical in addressing these gaps, but technical, legal, and ethical challenges arise. On the technical side, approaches for integrating complex and messy data are needed. Data and design imperfections like selection bias, missing data, and confounding must be addressed. In terms of legal and ethical challenges, while AI has the potential to aid in leveraging patient data to make clinical care decisions, we also risk exacerbating existing disparities. Organizations implementing AI solutions must carefully consider how they can improve care for all and reduce inequities.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Identifying Regulatory Elements via Deep Learning

引用

Biomedical Data Science 1000年第1期3卷 315-338页

作者： Mira Barshai Eitamar Tripto Yaron Orenstein 1School of Electrical and Computer Engineering Ben-Gurion University of the Negev Beer-Sheva 8410501 Israel email: yaronore@bgu.ac.il 2Department of Biomedical Engineering Ben-Gurion University of the Negev Beer-Sheva 8410501 Israel

Deep neural networks have been revolutionizing the field of machine learning for the past several years. They have been applied with great success in many domains of the biomedical data sciences and are outperforming extant methods by a large margin. The ability of deep neural networks to pick up local image features and model the interactions between them makes them highly applicable to regulatory genomics. Instead of an image, the networks analyze DNA and RNA sequences and additional epigenomic data. In this review, we survey the successes of deep learning in the field of regulatory genomics. We first describe the fundamental building blocks of deep neural networks, popular architectures used in regulatory genomics, and their training process on molecular sequence data. We then review several key methods in different gene regulation domains. We start with the pioneering method DeepBind and its successors, which were developed to predict protein–DNA binding. We then review methods developed to predict and model epigenetic information, such as histone marks and nucleosome occupancy. Following epigenomics, we review methods to predict protein–RNA binding with its unique challenge of incorporating RNA structure information. Finally, we provide our overall view of the strengths and weaknesses of deep neural networks and prospects for future developments.

关键词： deep learning regulatory genomics gene regulation motif finding

来源：评论

学校读者我要写书评

暂无评论

Synthesis for Robots: Guarantees and Feedback for Robot Behavior

引用

Control, Robotics, and Autonomous Systems 1000年第1期1卷 211-236页

作者： Hadas Kress-Gazit Morteza Lahijanian Vasumathi Raman 1Sibley School of Mechanical and Aerospace Engineering Cornell University Ithaca New York 14853 USA email: hadaskg@cornell.edu 2Department of Computer Science University of Oxford Oxford OX1 3QD United Kingdom email: morteza.lahijanian@cs.ox.ac.uk 3San Francisco California 94103 USA email: vasu@cds.caltech.edu

Robot control for tasks such as moving around obstacles or grasping objects has advanced significantly in the last few decades. However, controlling robots to perform complex tasks is still accomplished largely by highly trained programmers in a manual, time-consuming, and error-prone process that is typically validated only through extensive testing. Formal methods are mathematical techniques for reasoning about systems, their requirements, and their guarantees. Formal synthesis for robotics refers to frameworks for specifying tasks in a mathematically precise language and automatically transforming these specifications into correct-by-construction robot controllers or into a proof that the task cannot be done. Synthesis allows users to reason about the task specification rather than its implementation, reduces implementation error, and provides behavioral guarantees for the resulting controller. This article reviews the current state of formal synthesis for robotics and surveys the landscape of abstractions, specifications, and synthesis algorithms that enable it.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exposed! A Survey of Attacks on Private Data

引用

Statistics and Its Application 1000年第1期4卷 61-84页

作者： Cynthia Dwork Adam Smith Thomas Steinke Jonathan Ullman 1Microsoft Research Mountain View California 94043 email: dwork@*** 2Department of Computer Science and Engineering Pennsylvania State University State College Pennsylvania 16802 email: asmith@psu.edu 3John A. Paulson School of Engineering and Applied Sciences Harvard University Cambridge Massachusetts 02138 email: tsteinke@seas.harvard.edu 4College of Computer and Information Science Northeastern University Boston Massachusetts 02115 email: jullman@ccs.neu.edu

Privacy-preserving statistical data analysis addresses the general question of protecting privacy when publicly releasing information about a sensitive dataset. A privacy attack takes seemingly innocuous released information and uses it to discern the private details of individuals, thus demonstrating that such information compromises privacy. For example, re-identification attacks have shown that it is easy to link supposedly de-identified records to the identity of the individual concerned. This survey focuses on attacking aggregate data, such as statistics about how many individuals have a certain disease, genetic trait, or combination thereof. We consider two types of attacks: reconstruction attacks, which approximately determine a sensitive feature of all the individuals covered by the dataset, and tracing attacks, which determine whether or not a target individual's data are included in the dataset. We also discuss techniques from the differential privacy literature for releasing approximate aggregate statistics while provably thwarting any privacy attack.

关键词： privacy privacy attacks re-identification reconstruction attacks tracing attacks differential privacy

来源：评论

学校读者我要写书评

暂无评论

Control Across Scales by Positive and Negative Feedback

引用

Control, Robotics, and Autonomous Systems 1000年第1期2卷 89-113页

作者： R. Sepulchre G. Drion A. Franci 1Department of Engineering University of Cambridge Cambridge CB2 1PZ United Kingdom email: r.sepulchre@eng.cam.ac.uk 2Department of Electrical Engineering and Computer Science University of Liège 4000 Liège Belgium email: gdrion@uliege.be 3Department of Mathematics Faculty of Sciences National Autonomous University of Mexico Mexico City 04510 Mexico email: afranci@ciencias.unam.mx

Feedback is a key element of regulation, as it shapes the sensitivity of a process to its environment. Positive feedback upregulates, and negative feedback downregulates. Many regulatory processes involve a mixture of both, whether in nature or in engineering. This article revisits the mixed-feedback paradigm, with the aim of investigating control across scales. We propose that mixed feedback regulates excitability and that excitability plays a central role in multiscale neuronal signaling. We analyze this role in a multiscale network architecture inspired by neurophysiology. The nodal behavior defines a mesoscale that connects actuation at the microscale to regulation at the macroscale. We show that mixed-feedback nodal control provides regulatory principles at the network scale, with a nodal resolution. In this sense, the mixed-feedback paradigm is a control principle across scales.

关键词： feedback control excitability ultrasensitivity positive and negative feedback neuronal networks multiscale control

来源：评论

学校读者我要写书评

暂无评论

Computational Methods for Single-Cell RNA Sequencing

引用

Biomedical Data Science 1000年第1期3卷 339-364页

作者： Brian Hie Joshua Peters Sarah K. Nyquist Alex K. Shalek Bonnie Berger Bryan D. Bryson 1Computer Science and Artificial Intelligence Laboratory (CSAIL) Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA email: bab@mit.edu 3Ragon Institute of MGH MIT and Harvard Cambridge Massachusetts 02139 USA 2Department of Biological Engineering Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA email: bryand@mit.edu 4Program in Computational and Systems Biology Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA 5Department of Chemistry Institute for Medical Engineering & Science (IMES) and Koch Institute for Integrative Cancer Research Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA 6Department of Mathematics Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA

Single-cell RNA sequencing (scRNA-seq) has provided a high-dimensional catalog of millions of cells across species and diseases. These data have spurred the development of hundreds of computational tools to derive novel biological insights. Here, we outline the components of scRNA-seq analytical pipelines and the computational methods that underlie these steps. We describe available methods, highlight well-executed benchmarking studies, and identify opportunities for additional benchmarking studies and computational methods. As the biochemical approaches for single-cell omics advance, we propose coupled development of robust analytical pipelines suited for the challenges that new data present and principled selection of analytical methods that are suited for the biological questions to be addressed.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Harnessing Protease Activity to Improve Cancer Care

引用

Cancer Biology 1000年第1期2卷 353-376页

作者： Jaideep S. Dudani Andrew D. Warren Sangeeta N. Bhatia 2Department of Biological Engineering Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA 1Koch Institute for Integrative Cancer Research Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA email: jdudani@mit.edu andrew.warren11@*** sbhatia@mit.edu 3Harvard–MIT Division of Health Sciences and Technology Institute for Medical Engineering and Science Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA 7Howard Hughes Medical Institute Cambridge Massachusetts 02139 USA 6Broad Institute of Massachusetts Institute of Technology and Harvard Cambridge Massachusetts 02139 USA 5Department of Medicine Brigham and Women's Hospital and Harvard Medical School Boston Massachusetts 02115 USA 4Department of Electrical Engineering and Computer Science Massachusetts Institute of Technology Cambridge Massachusetts 02139 USA

Proteolysis plays critical roles in normal and pathologic physiology; these enzymes are intricately involved in cancer progression and spread. Our understanding of protease function has advanced from nonspecific degrading enzymes to a modern appreciation of their diverse roles in posttranslational modification and signaling in a complex microenvironment. This new understanding has led to next-generation diagnostics and therapeutics that exploit protease activity in cancer. For diagnostics, protease activity may be measured as a biomarker of cancer, with wide-ranging utility from early detection to monitoring therapeutic response. Therapeutically, while broad inhibition of protease activity proved disappointing, new approaches that more specifically modulate proteases in concert with secondary targets might enable potent combination therapies. In addition, clinical evaluation is underway for tools that leverage protease activity to activate therapeutics, ranging from imaging agents that monitor surgical margins to immunotherapies with improved specificity. Technologies that interact with, measure, or modulate proteases are poised to improve cancer management on diagnostic and therapeutic fronts to realize the promise of precision medicine.

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：