检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Lindner, David Shah, Rohin Abbeel, Pieter Dragan, Anca Department of Computer Science ETH Zurich Center for Human-Compatible AI UC Berkeley

Since reward functions are hard to specify, recent work has focused on learning policies from human feedback. However, such approaches are impeded by the expense of acquiring such feedback. Recent work proposed that agents have access to a source of information that is effectively free: in any environment that humans have acted in, the state will already be optimized for human preferences, and thus an agent can extract information about what humans want from the state (Shah et al., 2019). Such learning is possible in principle, but requires simulating all possible past trajectories that could have led to the observed state. This is feasible in gridworlds, but how do we scale it to complex tasks? In this work, we show that by combining a learned feature encoder with learned inverse models, we can enable agents to simulate human actions backwards in time to infer what they must have done. The resulting algorithm is able to reproduce a specific skill in MuJoCo environments given a single state sampled from the optimal policy for that skill. Copyright © 2021, The Authors. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

A Numerical real time web tracking and scrapping strategy applied to analysing covid-19 datasets 7

A Numerical real time web tracking and scrapping strategy ap...

引用

7th Annual International Conference on Network and Information Systems for computers, ICNISC 2021

作者： Ez-Zahout, Abderrahmane Chakouk, Said Mitouilli, Salwa El Bouni, Mohamed Amine Mohammed v University in Rabat Faculty of Sciences Computer Sciences Department Morocco Mohammed v University in Rabat Faculty of Letters and Human Sciences Communication Department Morocco Al Akhawayn University in Ifrane School of Science and Engineering Morocco

ISBN: (纸本)9781665402323

With the huge amount of data collected from the web, it is hard to manually analyze and extract useful incites from tables, matrices, or rows of data. Therefore, we need a way to represent these data (maps or graphs) to analyze/interpret them. Data visualization makes it easier to identify trends, patterns, and outliers with large datasets. Data Analytics is used to analyze raw data to make conclusions about the information. In this paper, we tackled web scrapping strategies to get COVID-19 data from the web, and then we went through some data analysis and exploration using EDA. © 2021 IEEE.

关键词： Data visualization

来源：评论

学校读者我要写书评

暂无评论

A Detailed Review and Study of Implementation of Security System Design for Machine Learning-Based Software Systems: Issues And Challenges

A Detailed Review and Study of Implementation of Security Sy...

引用

Advance Computing and Innovative Technologies in Engineering (ICACITE), International Conference on

作者： Melanie Lourens Vikas Tripathi Joel Alanya-Beltran Shaik Vaseem Akram Bandi Bhaskar Firos A Department of Human Resources Management Durban University of Technology Berea Durban South Africa Department of Computer Science & Engineering Graphic Era Deemed to be University Dehradun Uttarakhand India Universidad Tecnológica del Perú Perú Uttaranchal Institute of Technology Uttaranchal University School of CS& AI SR University Warangal Telangana Department of Computer Science and Engineering Rajiv Gandhi University (A Central University) Doimukh Arunachal Pradesh India

The rapidly growing technology specially in the field of software, Machine Learning (ML) has played an important role in a range of tasks, including voice, video, and computer vision. It is currently being utilised in software systems to automate the crucial processes more and more. Machine learning-based modern software systems (MLBSS) are currently difficult to build safely, which will severely limit the uses in security and safety-critical domains. Recently, majority of articles are published and still research work is going on the safety problems for ML and Deep Learning (DL), which place a strong prominence on the models and data both, adversaries’ threats have been taken into consideration. In this paper, we address the prospect that system bugs or external adversarial assaults might lead to security vulnerabilities for machine learning-based software systems, and we propose that safe development techniques ought to be applied throughout the whole lifecycle. We conclude by providing a thorough study of the security for MLBSS, which includes a comprehensive analysis based on a review of the structure of three distinctive features in terms of security issues. The entire state-of-the-art for MLBSS secure development is also provided.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Privacy-Preserving Graph Convolutional Networks for Text Classification

arXiv

引用

arXiv 2021年

作者： Igamberdiev, Timour Habernal, Ivan Trustworthy Human Language Technologies Department of Computer Science Technical University of Darmstadt Germany

Graph convolutional networks (GCNs) are a powerful architecture for representation learning on documents that naturally occur as graphs, e.g., citation or social networks. However, sensitive personal information, such as documents with people's profiles or relationships as edges, are prone to privacy leaks, as the trained model might reveal the original input. Although differential privacy (DP) offers a well-founded privacy-preserving framework, GCNs pose theoretical and practical challenges due to their training specifics. We address these challenges by adapting differentially-private gradient-based training to GCNs and conduct experiments using two optimizers on five NLP datasets in two languages. We propose a simple yet efficient method based on random graph splits that not only improves the baseline privacy bounds by a factor of 2.7 while retaining competitive F1scores, but also provides strong privacy guarantees of ϵ = 1:0. We show that, under certain modeling choices, privacy-preserving GCNs perform up to 90% of their non-private variants, while formally guaranteeing strong privacy measures. © 2021, CC BY-SA.

关键词： Classification (of information)

来源：评论

学校读者我要写书评

暂无评论

The Role of AI in Healthcare Policy Development and Management

The Role of AI in Healthcare Policy Development and Manageme...

引用

Artificial Intelligence for Innovations in Healthcare Industries (ICAIIHI), International Conference on

作者： Sapna Yadav Pravin Mane Virendra Kumar Swarnkar Shital kumar Rawandale Alaknanda S. Patil Kiran S. Katke Niyati Bhat Entrepreneurship Mindset Curriculum State Council of Educational Research and Training Delhi Bharati Vidyapeeth (Deemed to be University) Institute of Management and Entrepreneurship Development Pune Department of Computer Science & Engineering Bharti Vishwavidyalaya Durg Chhattisgarh Dean Industry Institute Interaction PCET's Pimpri Chinchwad College of Engineering Pune Maharashtra India E&TC department JSPM NTC Pune Department of Computer Engineering Sanjivani College of engineering Kopergaon Ahmednagar Software Developer Master of Science california state university long beach CA USA

This study looks at how artificial intelligence (AI) is incorporated into healthcare policy, with an emphasis on how it affects administrative procedures, treatment optimization, especially clinical choice support. Using a deductive methodology and interpretive thinking as a philosophy, an exploratory design with secondary data collecting was used. In the present-day environment, AI is being used more frequently in healthcare legislation, especially within medical settings. Clinical decision-making systems driven by AI have shown to offer a great promise for lowering diagnostic mistakes and increasing treatment precision. Improving patient outcomes through individualized AI-driven treatments for treatment improvement appears promising. But there are issues with algorithm prejudices as well as information privacy, among other ethical and legal issues. Establishing thorough guidelines, ongoing education, and auditing procedures are among the suggestions that are made. In order to guarantee fair and efficient integration of health care legislation, future research should prioritize longitudinal research and improving artificial intelligence algorithms

关键词：

来源：评论

学校读者我要写书评

暂无评论

Efficient Utilization of Large Pre-Trained Models for Low Resource ASR

arXiv

引用

arXiv 2022年

作者： Vieting, Peter Lüscher, Christoph Dierkes, Julian Schlüter, Ralf Ney, Hermann Human Language Technology and Pattern Recognition Group Computer Science Department RWTH Aachen University Aachen52074 Germany AppTek GmbH Aachen52062 Germany

Unsupervised representation learning has recently helped automatic speech recognition (ASR) to tackle tasks with limited labeled data. Following this, hardware limitations and applications give rise to the question how to take advantage of large pre-trained models efficiently and reduce their complexity. In this work, we study a challenging low resource conversational telephony speech corpus from the medical domain in Vietnamese and German. We show the benefits of using unsupervised techniques beyond simple fine-tuning of large pre-trained models, discuss how to adapt them to a practical telephony task including bandwidth transfer and investigate different data conditions for pre-training and fine-tuning. We outperform the project baselines by 22% relative using pre-training techniques. Further gains of 29% can be achieved by refinements of architecture and training and 6% by adding 0.8 h of in-domain adaptation data. Copyright © 2022, The Authors. All rights reserved.

关键词： Speech recognition

来源：评论

学校读者我要写书评

暂无评论

TraSeTR: Track-to-segment transformer with contrastive query for instance-level instrument segmentation in robotic surgery

arXiv

引用

arXiv 2022年

作者： Zhao, Zixu Jin, Yueming Heng, Pheng-Ann Department Of Computer Science And Engineering The Chinese University Of Hong Kong Hong Kong Guangdong-Hong Kong-Macao Joint Laboratory Of Human-Machine Intelligence-Synergy Systems Shenzhen Institute Of Advanced Technology Chinese Academy Of Sciences China Department Of Computer Science University College London United Kingdom

Surgical instrument segmentation - in general a pixel classification task - is fundamentally crucial for promoting cognitive intelligence in robot-assisted surgery (RAS). However, previous methods are struggling with discriminating instrument types and instances. To address above issues, we explore a mask classification paradigm that produces per-segment predictions. We propose TraSeTR, a novel Track-to-Segment Transformer that wisely exploits tracking cues to assist surgical instrument segmentation. TraSeTR jointly reasons about the instrument type, location, and identity with instance-level predictions i.e., a set of class-bbox-mask pairs, by decoding query embeddings. Specifically, we introduce the prior query that encoded with previous temporal knowledge, to transfer tracking signals to current instances via identity matching. A contrastive query learning strategy is further applied to reshape the query feature space, which greatly alleviates the tracking difficulty caused by large temporal variations. The effectiveness of our method is demonstrated with state-of-the-art instrument type segmentation results on three public datasets, including two RAS benchmarks from EndoVis Challenges and one cataract surgery dataset CaDIS. © 2022, CC BY-NC-SA.

关键词： Surgical equipment

来源：评论

学校读者我要写书评

暂无评论

What if it speaks like it was from the village? Effects of a Robot speaking in Regional Language Variations on Users’ Evaluations

What if it speaks like it was from the village? Effects of a...

引用

IEEE International Workshop on Robot and human Communication (ROMAN)

作者： Birgit Lugrin Elisabeth Ströle David Obremski Frank Schwab Benjamin Lange Human-Computer Interaction Institute of Computer Science University of Würzburg Media Psychology Instutute Human Computer Media University of Würzburg

ISBN: (数字)9781728160757

ISBN: (纸本)9781728160764

The present contribution investigates the effects of spoken language varieties, in particular non-standard / regional language compared to standard language (in our study: High German), in social robotics. Based on (media) psychological and sociolinguistic research, we assumed that a robot speaking in regional language (i.e., dialect and regional accent) would be considered less competent compared to the same robot speaking in standard language (H1). Contrarily, we assumed that regional language might enhance perceived social skills and likability of a robot, at least so when taking into account whether and how much the human observers making the evaluations talk in regional language themselves. More precisely, it was assumed that the more the study participants spoke in regional language, the better their ratings of the dialect-speaking robot on social skills and likeability would be (H2). We also investigated whether the robot's gender (male vs. female voice) would have an effect on the ratings (RQ). H1 received full, H2 limited empirical support by the data, while the robot's gender (RQ) turned out to be a mostly negligible factor. Based on our results, practical implications for robots speaking in regional language varieties are suggested.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Video Face Clustering with Self-Supervised Representation Learning

IEEE Transactions on Biometrics, Behavior, and Identity Scie...

引用

IEEE Transactions on Biometrics, Behavior, and Identity science 2020年第2期2卷 145-157页

作者： Sharma, Vivek Tapaswi, Makarand Saquib Sarfraz, M. Stiefelhagen, Rainer Department of Informatics Institute for Anthropomatics and Robotics Computer Vision for Human-Computer Interaction Lab Karlsruhe Institute of Technology Karlsruhe76131 Germany University of Toronto TorontoONM5S 3G4 Canada Massachusetts Institute of Technology CambridgeMA02139 United States Vector Institute TorontoON Canada Inria Paris75012 France

Characters are a key component of understanding the story conveyed in TV series and movies. With the rise of advanced deep face models, identifying face images may seem like a solved problem. However, as face detectors get better, clustering and identification need to be revisited to address increasing diversity in facial appearance. In this paper, we propose unsupervised methods for feature refinement with application to video face clustering. Our emphasis is on distilling the essential information, identity, from the representations obtained using deep pre-trained face networks. We propose a self-supervised Siamese network that can be trained without the need for video/track based supervision, that can also be applied to image collections. We evaluate our methods on three video face clustering datasets. Thorough experiments including generalization studies show that our methods outperform current state-of-the-art methods on all datasets. The datasets and code are available at https://***/vivoutlaw/SSIAM. © 2019 IEEE.

关键词： computer vision

来源：评论

学校读者我要写书评

暂无评论

A systematic study on the role of SentiWordNet in opinion mining

引用

Frontiers of computer science 2021年第4期15卷 165-183页

作者： Mujtaba HUSNAIN Malik Muhammad Saad MISSEN Nadeem AKHTAR Mickael COUSTATY Shahzad MUMTAZ V.B.Surya PRASATH Department of IT Faculty of ComputingThe Islamia University of BahawalpurBahawalpur 63100Pakistan Laboratoire Informatique Image et Interaction(L3i)Facultés des Sciences et TechnologiesUniversity of La RochelleLa Rochelle 17000France Division of Biomedical Informatics Cincinnati Children’s Hospital Medical CenterCincinnati 45229USA Department of Pediatrics University of Cincinnati College of MedicineCincinnati 45229USA Department of Biomedical Informatics College of MedicineUniversity of CincinnatiCincinnati 45267USA Department of Electrical Engineering and Computer Science University of CincinnatiCincinnati 45221USA

Sentiment lexicons(SL)(aka lexical resources)are the repositories of one or several dictionaries that consist of known and precompiled sentiment *** lexicons play an important role in performing several different opinion mining *** efficacy of the lexicon-based approaches in performing opinion mining(OM)tasks solely depends on selecting an appropriate opinion lexicon to analyze the ***,one has to explore the available sentiment lexicons and then select the most suitable *** available resources,SentiWordNet(SWN)is the most widely used lexicon to perform tasks related to opinion *** SWN,each synset of WordNet is being assigned the three sentiment numerical scores;positive,negative and objective that are calculated using by a set of *** this paper,a detailed and comprehensive review of the work related to opinion mining using Senti-WordNet is provided in a very distinctive *** survey will be useful for the researchers contributing to the field of opinion *** features make our contribution worthwhile and unique among the reviews of similar kind:(i)our review classifies the existing literature with respect to opinion mining tasks and subtasks(ii)it covers a very different outlook of the opinion mining field by providing in-depth discussions of the existing works at different granularity levels(word,sentences,document,aspect,clause,and concept levels)(iii)this state-ofart review covers each article in the following dimensions:the designated task performed,granularity level of the task completed,results obtained,and feature dimensions,and(iv)lastly it concludes the summary of the related articles according to the granularity levels,publishing years,related tasks(or subtasks),and types of classifiers *** the end,major challenges and tasks related to lexicon-based approaches towards opinion mining are also discussed.

关键词： opinion mining lexical based resources Senti-WordNet opinion strength

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：