检索结果-内蒙古大学图书馆

arXiv 2019年

作者： Vig, Jesse Belinkov, Yonatan Palo Alto Research Center Machine Learning and Data Science Group Interaction and Analytics Lab Palo AltoCA United States Harvard John A. Paulson School of Engineering and Applied Sciences MIT Computer Science and Artificial Intelligence Laboratory CambridgeMA United States

The Transformer is a fully attention-based alternative to recurrent networks that has achieved state-of-the-art results across a range of NLP tasks. In this paper, we analyze the structure of attention in a Transformer language model, the GPT-2 small pretrained model. We visualize attention for individual instances and analyze the interaction between attention and syntax over a large corpus. We find that attention targets different parts of speech at different layer depths within the model, and that attention aligns with dependency relations most strongly in the middle layers. We also find that the deepest layers of the model capture the most distant relationships. Finally, we extract exemplar sentences that reveal highly specific patterns targeted by particular attention heads. Copyright © 2019, The Authors. All rights reserved.

关键词： Recurrent neural networks

来源：评论

学校读者我要写书评

暂无评论

Early Stage Kidney Chronic Diseases Diagnosis using Feature Fusion and Deep Learning Algorithms

Early Stage Kidney Chronic Diseases Diagnosis using Feature ...

引用

International Conference on Developments in eSystems Engineering (DESE)

作者： Tanzila Saba Muhamamd Mujahid Noor Ayesha Mahyar Kolivand Artificial Intelligence & Data Analytics Lab CCIS Prince Sultan University Riyadh Saudi Arabia School of Clinical Medicine Zhengzhou University Zhengzhou Henan RP China Department of Medicine University of Liverpool Liverpool UK School of Computing and Digital Technologies Staffordshire University Staffordshire UK

ISBN: (数字)9798350368697

ISBN: (纸本)9798350368703

Chronic Kidney Disease (CKD) is a global health issue due to pre-existing conditions like anemia, diabetes, and hypertension. Early warning signals are commonly neglected, complicating diagnosis and treatment. This study uses high-performance data mining techniques to provide self-analysis and prognostic tools for medical professionals. Clinical decision support using domain experts and specialists aids CKD prognosis and rehabilitation. We use advanced data mining to evaluate demographic characteristics, test findings, and patient medical information to uncover insights and patterns. We used a modified artificial neural network to extract characteristics from a dataset and long-term and short-term memory to classify chronic kidney disease stages. LSTM algorithms improve the accuracy of CKD diagnosis and stage prediction. Initial validation tests demonstrate a 92.23% accuracy gain over traditional methods. The binary classification AUC was 99.95% and the multi-class classification was 94.13% with the proposed technique. LSTM algorithms use temporal dynamics in patient data to enhance diagnostic accuracy, early diagnosis, therapy customization, and healthcare costs. We proposed a novel CKD diagnostic and prognosis paradigm and test multiple classification algorithms for disease diagnosis and staging. Multiple methods improve reliability and account for CKD diagnostic peculiarities.

关键词： Support vector machines Accuracy Heuristic algorithms Prediction algorithms Chronic kidney disease Classification algorithms data mining Medical diagnosis Prognostics and health management Medical diagnostic imaging

来源：评论

学校读者我要写书评

暂无评论

Rethinking performance estimation in neural architecture search

arXiv

引用

arXiv 2020年

作者： Zheng, Xiawu Ji, Rongrong Wang, Qiang Ye, Qixiang Li, Zhenguo Tian, Yonghong Tian, Qi Media Analytics and Computing Lab Department of Artificial Intelligence School of Informatics Xiamen University 361005 China National Institute for Data Science in Health and Medicine Xiamen University Peng Cheng Laboratory Shenzhen China University of Chinese Academy of Sciences China Noah's Ark Lab Huawei Technologies Peking University

Neural architecture search (NAS) remains a challenging problem, which is attributed to the indispensable and time-consuming component of performance estimation (PE). In this paper, we provide a novel yet systematic rethinking of PE in a resource constrained regime, termed budgeted PE (BPE), which precisely and effectively estimates the performance of an architecture sampled from an architecture space. Since searching an optimal BPE is extremely time-consuming as it requires to train a large number of networks for evaluation, we propose a Minimum Importance Pruning (MIP) approach. Given a dataset and a BPE search space, MIP estimates the importance of hyper-parameters using random forest and subsequently prunes the minimum one from the next iteration. In this way, MIP effectively prunes less important hyper-parameters to allocate more computational resource on more important ones, thus achieving an effective exploration. By combining BPE with various search algorithms including reinforcement learning, evolution algorithm, random search, and differentiable architecture search, we achieve 1, 000× of NAS speed up with a negligible performance drop comparing to the SOTA.1. Copyright © 2020, 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

WebPPG: Feasibility and Usability of Self-Performed, Browser-Based Smartphone Photoplethysmography

WebPPG: Feasibility and Usability of Self-Performed, Browser...

引用

Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBC)

作者： Michael Nissen Madeleine Flaucher Katharina M. Jaeger Hanna Huebner Nina Danzberger Adriana Titzmann Constanza A. Pontones Peter A. Fasching Bjoern M. Eskofier Heike Leutheuser Department Artificial Intelligence in Biomedical Engineering (AIBE) Machine Learning and Data Analytics Lab (MaD Lab) Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen Germany Department of Gynecology and Obstetrics Erlangen University Hospital Friedrich-Alexander-Universität Erlangen-Nürnberg (FAU) Erlangen Germany

Smartphones enable and facilitate biomedical studies as they allow the recording of various biomedical signals, including photoplethysmograms (PPG). However, user engagement rates in mobile health studies are reduced when an application (app) needs to be installed. This could be alleviated by using installation-free web apps. We evaluate the feasibility of browser-based PPG recording, conducting the first usability study on smartphone-based PPG. We present an at-home study using a web app and library for PPG recording using the rear camera and flash. The underlying library is freely made available to researchers. 25 Android users participated, using their own smartphones. The study consisted of a demographic and anamnestic questionnaire, the signal recording itself (60 s), and a consecutive usability questionnaire. After filtering, heart rate was extracted (14/17 successful), signal-to-noise ratios assessed (0.64 ± 0.50 dB, mean ± standard deviation), and quality was visually inspected (12/17 usable for diagnosis). Recording was not supported in 9 cases. This was due to the browser’s insufficient support for the flash light API. The app received a System Usability Scale score of 82 ± 9, which is above the 90th percentile. Overall, browser flash light support is the main limiting factor for broad device support. Thus, browser-based PPG is not yet widely applicable, although most participants feel comfortable with the recording itself. The utilization of the user-facing camera might represent a more promising approach. This study contributes to the development of low-barrier, user-friendly, installation-free smartphone signal acquisition. This enables profound, comprehensive data collection for research and clinical *** relevance— WebPPG offers low-barrier remote diagnostic capabilities without the need for app installation

关键词：

来源：评论

学校读者我要写书评

暂无评论

Semantics-Guided Contrastive Network for Zero-Shot Object Detection

arXiv

引用

arXiv 2021年

作者： Yan, Caixia Chang, Xiaojun Luo, Minnan Liu, Huan Zhang, Xiaoqin Zheng, Qinghua The Ministry of Education Key Lab of Intelligent Networks and Network Security National Engineering Lab for Big Data Analytics School of Electronic and Information Engineering Xi'an Jiaotong University Shaanxi710049 China The ReLER Lab AAAI Faculty of Engineering and Information Technology University of Technology Sydney Australia The College of Computer Science and Artificial Intelligence Wenzhou University China

Zero-shot object detection (ZSD), the task that extends conventional detection models to detecting objects from unseen categories, has emerged as a new challenge in computer vision. Most existing approaches tackle the ZSD task with a strict mapping-transfer strategy that may lead to suboptimal ZSD results: 1) the learning process of these models neglects the available semantic information on unseen classes, which can easily bias towards the seen categories;2) the original visual feature space is not well-structured for the ZSD task due to the lack of discriminative information. To address these issues, we develop a novel Semantics-Guided Contrastive Network for ZSD, named ContrastZSD, a detection framework that first brings contrastive learning mechanism into the realm of zero-shot detection. Particularly, ContrastZSD incorporates two semantics-guided contrastive learning subnets that contrast between region-category and region-region pairs respectively. The pairwise contrastive tasks take advantage of supervision signals derived from both the ground truth label and class similarity information. By performing supervised contrastive learning over those explicit semantic supervision, the model can learn more knowledge about unseen categories to avoid the bias problem to seen concepts, while optimizing the visual data structure to be more discriminative for better visual-semantic alignment. Extensive experiments are conducted on two popular benchmarks for ZSD, i.e., PASCAL VOC and MS COCO. Results show that our method outperforms the previous state-of-the-art on both ZSD and generalized ZSD tasks. © 2021, CC BY.

关键词： Object detection

来源：评论

学校读者我要写书评

暂无评论

Human Brain Stroke Prediction using Machine Learning Methods with Synthetic Minority Oversampling Approach

Human Brain Stroke Prediction using Machine Learning Methods...

引用

Women in data Science at Prince Sultan University (WiDS PSU), International Conference of

作者： Muhammad Mujahid Noor Ayesha Ahmad Taher Azar Tanzila Saba Zeeshan Haider Artificial Intelligence and Data Analytics (AIDA) Lab CCIS Prince Sultan University Riyadh School of Clinical Medicine Zhengzhou University Zhengzhou China College of Computer and Information Sciences Prince Sultan University Riyadh Saudi Arabia Automated Systems and Soft Computing Lab (ASSCL) Prince Sultan University Riyadh Saudi Arabia Faculty of Computers and Artificial Intelligence Benha University Benha Egypt Automated Systems and Soft Computing Lab (ASSCL) Prince Sultan University Riyadh Saudi Arabia

ISBN: (数字)9798350395839

ISBN: (纸本)9798350395846

The brain is considered to be the principal anatomical component of the human body. A stroke is a medical condition defined by cerebral injury that occurs as a consequence of blood vessel rupture within the brain. Symptoms might become apparent when there is an interruption in the brain circulation of blood and other vital chemicals. Stroke is widely acknowledged as a significant medical disorder because to its elevated mortality rates, enduring cognitive impairments, and concomitant comorbidities. The timely recognition of the numerous prodromal manifestations linked to a stroke has the capacity to attenuate its magnitude. The primary objective of this study is to utilize machine learning methodologies in order to predict the likelihood of an early-stage occurrence of a cerebral stroke. In order to assess the efficacy of the algorithm, a dataset for stroke prediction was acquired via the Kaggle platform. Various preprocessing techniques are applied to normalize and clean the dataset for stroke prediction. This research utilized many classification models including Random Forest, Ada Boost, LightGBM,, Decision Tree (DT), Logistic Regression (LR),and Support vector machine (SVM) to perform classification tasks. The Random Forest classifier has a higher level of classification accuracy in comparison to alternative machine learning classifiers, obtaining an impressive accuracy rate of 99.45%. The results of the investigation revealed that machine learning approaches demonstrated a higher level of performance in comparison to alternative methods.

关键词： Support vector machines Deep learning Accuracy Machine learning algorithms Computed tomography Stroke (medical condition) Prediction algorithms

来源：评论

学校读者我要写书评

暂无评论

NC-ALG: Graph-Based Active Learning Under Noisy Crowd

NC-ALG: Graph-Based Active Learning Under Noisy Crowd

引用

International Conference on data Engineering

作者： Wentao Zhang Yexin Wang Zhenbang You Yang Li Gang Cao Zhi Yang Bin Cui Center for Machine Learning Research Peking University Institute of Advanced Algorithms Research Shanghai National Engineering Labratory for Big Data Analytics and Applications Key Lab of High Confidence Software Technologies Peking University Department of Data Platform TEG Tencent Inc. Beijing Academy of Artificial Intelligence Institute of Computational Social Science Peking University Qingdao

ISBN: (数字)9798350317152

ISBN: (纸本)9798350317169

Graph Neural Networks (GNNs) have achieved great success in various data mining tasks but they heavily rely on a large number of annotated nodes, requiring considerable human efforts. Despite the effectiveness of existing GNN-based Active Learning (AL) methods, they assume that the annotated labels are always correct, which is contradictory to the error-prone labeling process in a practical crowdsourcing environment. Besides, due to this impractical assumption, existing works only focus on optimizing the node selection in AL but neglect optimizing the labeling process. Therefore, we present NC-ALG, the first GNN-based AL framework that optimizes both the node selection and node labeling process under a noisy crowd. For node selection, NC-ALG introduces a new measurement to model influence reliability and an effective influence maximization objective to select nodes. For node labeling, NC-ALG significantly reduces the labeling cost by considering the model-predicted labels and the labels of mirror nodes. To the best of our knowledge, this is the first attempt to consider GNN-based AL under the practical noisy crowd. Empirical studies on public datasets demonstrate that NC-ALG significantly outperforms existing methods in terms labeling efficiency. Notably, it only takes NC-ALG one-third of the labeling budget that the competitive baseline GRAIN needs to achieve an accuracy of 70.7 % on PubMed.

关键词： Crowdsourcing Costs Noise Predictive models Graph neural networks labeling Noise measurement

来源：评论

学校读者我要写书评

暂无评论

Automated Long-Term Monitoring of Stereotypical Movement in Polar Bears Under Human Care Using Machine Learning

SSRN

引用

SSRN 2024年

作者： Zuerl, Matthias Stoll, Philip Brehm, Ingrid Sueskind, Jonas Raab, René Petermann, Jan Zanca, Dario Simon, Ralph von Fersen, Lorenzo Eskofier, Bjoern Machine Learning and Data Analytics Lab Department Artificial Intelligence in Biomedical Engineering Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen91052 Germany Animal Physiology Department of Biology Friedrich-Alexander-Universität Erlangen-Nürnberg Erlangen91058 Germany Behavioral Ecology and Conservation Lab Nuremberg Zoo Nuremberg90480 Germany

The welfare of captive animals is often assessed by observing behaviours indicative of stress or discomfort, such as stereotypical behaviour (SB), which often shows as repetitive, invariant *** behaviour monitoring methods, however, are labour-intensive and subject to observer *** study presents an innovative automated approach utilising computer vision and machine learning to non-invasively detect and analyse SB in captive animals, exemplified by a longitudinal study of two polar *** designed an animal tracking framework to localise and identify individual animals in the *** determining their position on the enclosure map via homographic transformation, we cleaned the resulting trajectories using a particle ***, we classified the trajectory patterns as SB or normal behaviour using a lightweight random forest approach with an accuracy of 94.9 %.The system not only allows for continuous, objective monitoring of animal behaviours but also provides insights into seasonal variations in SB, illustrating its potential for improving animal welfare in zoological *** work advances the field of animal welfare research by introducing a scalable, efficient method for the long-term, automated monitoring of stereotypical behaviour, paving the way for its application across different species and settings. © 2024, The Authors. All rights reserved.

关键词： Enclosures

来源：评论

学校读者我要写书评

暂无评论

Comparative Evaluation of Attribute-Enabled Supervised Classification in Predicting the Air Quality

引用

Spatial Information Research 2023年第4期31卷 399-407页

作者： Subbulakshmi, P. Vimal, S. Robinson, Y. Harold Verma, Amit Nayak, Janmenjoy School of computer science and Engineering VIT Chennai campus Tamilnadu India Data Analytics Lab Department of Artificial Intelligence and Data Science Ramco Institute of Technology Rajapalayam India Department of Computer Science and Engineering Francis Xavier Engineering College Tirunelveli India Department of Computer Science & Engineering University Centre for Research & Development Chandigarh University Gharuan Mohali India Department of Computer Science Maharaja Sriram Chandra Bhanja Deo University Odisha India

Air pollution demonstrates the appearance of toxins into the air which is blocking human prosperity and the earth. It will portray as potentially the riskiest threats that humanity anytime faced. It makes hurt animals, harvests to thwart these issues in transportation territories need to expect air quality from pollutions utilizing AI systems and IoT. Along these lines, air quality evaluation and assumption has become a huge target for human health factors and also affect internal organs related to respiratory. The accuracy of Air Pollution prediction has been involved with the machine learning techniques and the best accuracy model is identified. The air quality prediction dataset is used for identifying the meteorology air pollution data while the predicted model is involved the decision tree computation for predicting the toxin contents in the region, the Air quality indicator is used to assess the pollution level and monitoring the air quality. The performance analysis shows that the decision tree technique has produced the better results in the performance metrics of Accuracy, precision, recall, and F1-score with the minimized error values while the comparative evaluation of Attribute-enabled classification has identified the best technique for predicting the air quality.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Differentially Private Stochastic Convex Optimization in (Non)-Euclidean Space Revisited

arXiv

引用

arXiv 2023年

作者： Su, Jinyan Zhao, Changhong Wang, Di Mohamed bin Zayed University of Artificial Intelligence United Arab Emirates Department of Information Engineering The Chinese University of Hong Kong Hong Kong Provable Responsible AI and Data Analytics Lab India Computational Bioscience Research Center India Division of CEMSE King Abdullah University of Science and Technology Saudi Arabia

In this paper, we revisit the problem of Differentially Private Stochastic Convex Optimization (DP-SCO) in Euclidean and general dp spaces. Specifically, we focus on three settings that are still far from well understood: (1) DP-SCO over a constrained and bounded (convex) set in Euclidean space;(2) unconstrained DP-SCO in dp space;(3) DP-SCO with heavy-tailed data over a constrained and bounded set in dp space. For problem (1), for both convex and strongly convex loss functions, we propose methods whose outputs could achieve (expected) excess population risks that are only dependent on the Gaussian width of the constraint set, rather than the dimension of the space. Moreover, we also show the bound for strongly convex functions is optimal up to a logarithmic factor. For problems (2) and (3), we propose several novel algorithms and provide the first theoretical results for both cases when 1 © 2023, CC BY.

关键词： Convex optimization

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：