检索结果-内蒙古大学图书馆

arXiv 2022年

作者： ValizadehAslani, Taha Shi, Yiwen Wang, Jing Ren, Ping Zhang, Yi Hu, Meng Zhao, Liang Liang, Hualou Department of Electrical and Computer Engineering College of Engineering Drexel University PhiladelphiaPA United States College of Computing and Informatics Drexel University PhiladelphiaPA United States Office of Research and Standards Office of Generic Drugs Center for Drug Evaluation and Research United States Food and Drug Administration Silver SpringMD United States School of Biomedical Engineering Science and Health Systems Drexel University PhiladelphiaPA United States

Classification on long-tailed distributed data is a challenging problem, which suffers from serious class-imbalance and hence poor performance on tail classes with only a few samples. Owing to this paucity of samples, learning on the tail classes is especially challenging for the fine-tuning when transferring a pretrained model to a downstream task. In this work, we present a simple modification of standard fine-tuning to cope with these challenges. Specifically, we propose a two-stage fine-tuning: we first fine-tune the final layer of the pretrained model with class-balanced reweighting loss, and then we perform the standard fine-tuning. Our modification has several benefits: (1) it leverages pretrained representations by only fine-tuning a small portion of the model parameters while keeping the rest untouched;(2) it allows the model to learn an initial representation of the specific task;and importantly (3) it protects the learning of tail classes from being at a disadvantage during the model updating. We conduct extensive experiments on synthetic datasets of both two-class and multi-class tasks of text classification as well as a real-world application to ADME (i.e., absorption, distribution, metabolism, and excretion) semantic labeling. The experimental results show that the proposed two-stage fine-tuning outperforms both fine-tuning with conventional loss and fine-tuning with a reweighting loss on the above datasets. © 2022, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Online Knowledge Acquisition with the Selective Inherited Model

Online Knowledge Acquisition with the Selective Inherited Mo...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Xiaocong Du Shreyas Kolala Venkataramanaiah Zheng Li Jae-sun Seo Frank Liu Yu Cao School of Electrical Computer and Energy Engineering Arizona State University Tempe AZ USA School of Computing Informatics and Decision Systems Engineering Arizona State University Tempe AZ USA CSMD Oak Ridge National Lab Oak Ridge TN USA

ISBN: (数字)9781728169262

ISBN: (纸本)9781728169279

Continual learning, which updates machine learning models according to streaming data, is increasingly needed in the dynamic systems. Such a scenario requires both the preservation of previous knowledge, as well as the adaptation to new observations, with high computational and memory efficiency at the edge. Previous approaches attempt to learn the knowledge class by class from scratch, using either regularization based or memory replay-based methods. However, they still suffer from severe accuracy drop, a.k.a catastrophic forgetting, during this incremental process. Moreover, as the entire model is involved in each updating, their computation cost is too expensive for edge computing. In this work, we propose a novel brain- inspired paradigm named acquisitive learning (AL). Different from previous approaches that focus only on model adaptation, AL emphasizes the importance of both knowledge inheritance and acquisition: the model is first pre-trained and selected in the cloud (the selective inherited model) and then adapted to new knowledge (the acquisition). The quality of the inherited model is monitored by the landscape of the loss function, while the acquisition is realized by segmented training. The combination of both steps reduces accuracy drop by >10× on the CIFAR- 100 dataset. Furthermore, AL benefits edge computing with 5× reduction in latency per training image on FPGA prototype and 150× reduction in training FLOPs.

关键词： Adaptation models Computational modeling Task analysis Brain modeling Data models Training Knowledge acquisition

来源：评论

学校读者我要写书评

暂无评论

Fine-Tuning BERT for Automatic ADME Semantic Labeling in FDA Drug Labeling to Enhance Product-Specific Guidance Assessment

arXiv

引用

arXiv 2022年

作者： Shi, Yiwen Wang, Jing Ren, Ping ValizadehAslani, Taha Zhang, Yi Hu, Meng Liang, Hualou College of Computing and Informatics Drexel University PhiladelphiaPA United States Office of Research and Standards Office of Generic Drugs Center for Drug Evaluation and Research United States Food and Drug Administration Silver Spring MD United States Department of Electrical and Computer Engineering College of Engineering Drexel University PhiladelphiaPA United States School of Biomedical Engineering Science and Health Systems Drexel University PhiladelphiaPA United States

Product-specific guidances (PSGs) recommended by the United States Food and Drug Administration (FDA) are instrumental to promote and guide generic drug product development. To assess a PSG, the FDA assessor needs to take extensive time and effort to manually retrieve supportive drug information of absorption, distribution, metabolism, and excretion (ADME) from the reference listed drug labeling. In this work, we leveraged the state-of-the-art pre-trained language models to automatically label the ADME paragraphs in the pharmacokinetics section from the FDA-approved drug labeling to facilitate PSG assessment. We applied a transfer learning approach by fine-tuning the pre-trained Bidirectional Encoder Representations from Transformers (BERT) model to develop a novel application of ADME semantic labeling, which can automatically retrieve ADME paragraphs from drug labeling instead of manual work. We demonstrated that fine-tuning the pre-trained BERT model can outperform the conventional machine learning techniques, achieving up to 11.6% absolute F1 improvement. To our knowledge, we were the first to successfully apply BERT to solve the ADME semantic labeling task. We further assessed the relative contribution of pre-training and fine-tuning to the overall performance of the BERT model in the ADME semantic labeling task using a series of analysis methods such as attention similarity and layer-based ablations. Our analysis revealed that the information learned via fine-tuning is focused on task-specific knowledge in the top layers of the BERT, whereas the benefit from the pre-trained BERT model is from the bottom layers. © 2022, CC BY.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcement Learning-based Obstacle Avoidance for Robot Movement in Warehouse Environments

Deep Reinforcement Learning-based Obstacle Avoidance for Rob...

引用

IEEE International Conference on Civil Aviation Safety and Information Technology (ICCASIT)

作者： Keqin Li Jiajing Chen Dezhi Yu Tao Dajun Xinyu Qiu Jieting Lian Ryan Ji Shengyuan Zhang Zhenyu Wan Baiwei Sun Bo Hong Fanghao Ni Jiaqi Han Department of Computer Science AMA University Quezon Philippines Courant Institute New York University 251 Mercer St New York NY USA School of Information University of California Berkeley University Avenue and Oxford St Berkeley CA United States Master of Science in Mechanical Engineering School of Engineering Carnegie Mellon University 5000 Forbes Avenue Pittsburgh PA USA College of Engineering Northeastern University Seattle USA Master of Science in Database Technology School of Professional Studies New York University 50 West 4th Street New York NY USA College of Engineering Cornell University 616 Thurston Ave Ithaca NY US Master of Professional Science in Applied Statistics-Data Science College of Computing and Information Science Cornell University 616 Thurston Ave Ithaca NY USA Independent Researcher Beijing China Master of Computer Science Donald Bren School of Information and Computer Sciences University of California Irvine 510 Aldrich Hall Irvine CA USA School of Informatics Computing and Cyber Systems Northern Arizona University Flagstaff AZ USA Shanghai Institute of Technology Shanghai China

ISBN: (数字)9798350389418

ISBN: (纸本)9798350389425

This paper proposes a deep reinforcement learning based on the warehouse environment, the mobile robot obstacle avoidance Algorithm. Firstly, the value function network is improved based on the pedestrian interaction. We can learn to obtain the relative importance of the current state and the historical trajectory state as well as the joint impact on the robot’s obstacle avoidance strategy. Secondly, the reward function of reinforcement learning is designed based on the spatial behaviour of pedestrians, and the robot is punished for the state where the angle changes too much, so as to achieve the requirement of comfortable obstacle avoidance. Finally, the feasibility and effectiveness of the deep reinforcement learning-based mobile robot obstacle avoidance algorithm in the warehouse environment in the complex environment of the warehouse are verified through simulation experiments.

关键词： Legged locomotion Pedestrians Heuristic algorithms Robot sensing systems Feature extraction Trajectory Safety Data mining Collision avoidance Information technology

来源：评论

学校读者我要写书评

暂无评论

On-Road Information Extraction from Lidar Data Via Multiple Feature Maps 24

On-Road Information Extraction from Lidar Data Via Multiple ...

引用

2020 24th ISPRS Congress on Technical Commission I

作者： Wu, H. Xie, Z. Wen, C. Wang, C. Li, J. Fujian Key Laboratory of Sensing and Computing for Smart Cities School of Informatics Xiamen University Xiamen361005 China Departments of Geography and Environmental Management/Systems Design Engineering University of Waterloo WaterlooONN2L3G1 Canada

On-road information, including road boundaries, road markings, and road cracks, provides significant guidance or warning to all road users. Recently, the on-road information extraction from LiDAR data have been widely studied. However, for the LiDAR data with lower accuracy and higher noise, some detailed information, such as road boundary, is difficult to be extracted correctly. Furthermore, most of previous studies lack an exploration of efficiently extracting multiple on-road information from a single framework. In this paper, we propose a new framework that can simultaneously extract multiple on-road information from high accuracy LiDAR data and can also more robustly extract detailed road boundaries from low accuracy LiDAR data. First, we propose a Curb-Aware Ground Filter to extract ground points with rich curb structure features. Second, we transform the vertical density, elevation gradient and intensity features of the ground points into multiple feature maps and extract multiple on-road information from the feature maps by employing a semantic segmentation network. Experimental results on three datasets with different data accuracy demonstrate that our method outperforms other recent competitive methods. © 2020 Copernicus GmbH. All rights reserved.

关键词： Semantics

来源：评论

学校读者我要写书评

暂无评论

Application of Graph-Curvature Features in Auxiliary Diagnosis for Histopathological Image Identification of Gastric Cancer

SSRN

引用

SSRN 2023年

作者： He, Ruilin Li, Chen Yang, Xinyi Yang, Jinzhu Jiang, Tao Grzegorzek, Marcin Sun, Hongzan Microscopic Image and Medical Image Analysis Group College of Medicine and Biological Information Engineering Northeastern University Shenyang China Key Laboratory of Intelligent Computing in Medical Image Ministry of Education Northeastern University Liaoning Shenyang China School of Intelligent Medicine Chengdu University of Traditional Chinese Medicine Chengdu China International Joint Institute of Robotics and Intelligent Systems Chengdu University of Information Technology Chengdu China Institute for Medical Informatics University of Luebeck Luebeck Germany Department of Knowledge Engineering University of Economics in Katowice Katowice Poland Shengjing Hospital of China Medical University Shenyang China School of Computer Science and Engineering

Background: Histopathology diagnosis is often regarded as the final diagnostic method for malignant tumors, but it has some drawbacks. This paper explores a computer-aided diagnostic method that can identify benign and malignant gastric cancer with histopathology images. Method: This article obtains the most suitable process through multiple experiments, compared multiple methods and features for classification. Firstly, the U-net is applied to segment the image. Next, the nucleus is extracted from the segmented image and the Minimum Spanning Tree (MST) diagram structure is drawn. The third step is to extract the graph-curvature features of histopathology image according to the MST image. Finally, by inputting graph-curvature features into the classifier, the recognition results for benign or malignant can be obtained. Result: During the experiment, we use various methods for comparison. In the image segmentation stage, U-net, watershed algorithm and Otsu threshold segmentation methods are used respectively. Combined with multiple indicators, we find that the U-net method is the most suitable for segmentation of histopathology images. In the feature extraction stage, in addition to extracting graph-edge feature and graph-curvature feature, several basic image features are also extracted, including Red, Green, Blue feature, Gray-Level Co-occurrence Matrix feature, Histogram of Oriented Gradient feature, and Local Binary Pattern feature. In the classifier design stage, we experimented with various methods, such as Support vector machine (SVM), Random forest, Artificial Neural Network, K Nearest Neighbors, VGG-16 and Inception-V3. Through the comparison and analysis, the classification results with an accuracy of 98.57% can be obtained by inputting the graph-curvature feature into SVM classifier. Conclusion: This paper has created a unique feature, graph-curvature feature based on MST to represent and analyze histopathology images. This graph-based feature can be used

关键词： Local binary pattern

来源：评论

学校读者我要写书评

暂无评论

Fine-grained Domain Adaptive Crowd Counting via Point-derived Segmentation

arXiv

引用

arXiv 2021年

作者： Liu, Yongtuo Xu, Dan Ren, Sucheng Wu, Hanjie Cai, Hongmin He, Shengfeng School of Computer Science South China University of Technology China Institute of Informatics University of Amsterdam Netherlands Department of Computer Science and Engineering Hong Kong University of Science and Technology Hong Kong School of Computing and Information Systems Singapore Management University Singapore

Due to domain shift, a large performance drop is usually observed when a trained crowd counting model is deployed in the wild. While existing domain-adaptive crowd counting methods achieve promising results, they typically regard each crowd image as a whole and reduce domain discrepancies in a holistic manner, thus limiting further improvement of domain adaptation performance. To this end, we propose to untangle domain-invariant crowd and domain-specific background from crowd images and design a fine-grained domain adaption method for crowd counting. Specifically, to disentangle crowd from background, we propose to learn crowd segmentation from point-level crowd counting annotations in a weakly-supervised manner. Based on the derived segmentation, we design a crowd-aware domain adaptation mechanism consisting of two crowd-aware adaptation modules, i.e., Crowd Region Transfer (CRT) and Crowd Density Alignment (CDA). The CRT module is designed to guide crowd features transfer across domains beyond background distractions. The CDA module dedicates to regularising target-domain crowd density generation by its own crowd density distribution. Our method outperforms previous approaches consistently in the widely-used adaptation scenarios. © 2021, CC BY.

关键词： Image enhancement

来源：评论

学校读者我要写书评

暂无评论

Nature, Science, and PNAS – disciplinary profiles and impact

arXiv

引用

arXiv 2020年

作者： Milojević, Staša Center for Complex Networks and Systems Research Luddy School of Informatics Computing and Engineering Indiana University Bloomington United States

Nature, Science, and PNAS are the three most prestigious general-science journals, and Nature and Science are among the most influential journals overall, based on the journal Impact Factor (IF). In this paper we perform automatic classification of ~50,000 articles in these journals (published in the period 2005-2015) into 14 broad areas, to explore disciplinary profiles and to determine their field-specific IFs. We find that in all three journals the articles from Bioscience, Astronomy, and Geosciences are overrepresented, with other areas being under-represented, some of them severely. Discipline-specific IFs in these journals vary greatly, for example, between 18 and 46 for Nature. We find that the areas that have the highest disciplinary IFs are not the ones that contribute the most articles. We also find that publishing articles in these three journals brings prestige for articles in all areas, but at different levels, the least being for Astronomy. Comparing field-specific IFs of Nature, Science and PNAS to other top journals in six largest areas (Bioscience, Medicine, Geosciences, Physics, Astronomy, and Chemistry) these three journals are always among the top seven journals, with Nature being at the very top for all fields except in Medicine. Copyright © 2020, 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

Iterative embedding and reweighting of complex networks reveals community structure

arXiv

引用

arXiv 2024年

作者： Kovács, Bianka Kojaku, Sadamori Palla, Gergely Fortunato, Santo Department of Biological Physics Eötvös Loránd University Pázmány P. stny. 1/A BudapestH-1117 Hungary Luddy School of Informatics Computing and Engineering Indiana University 1015 East 11th Street BloomingtonIN47408 United States Department of Systems Science and Industrial Engineering SUNY Binghamton P.O. Box 6000 BinghamtonNY13902 United States Health Services Management Training Centre Semmelweis University Kútvölgyi út 2. BudapestH-1125 Hungary

Graph embeddings learn the structure of networks and represent it in low-dimensional vector spaces. Community structure is one of the features that are recognized and reproduced by embeddings. We show that an iterative procedure, in which a graph is repeatedly embedded and its links are reweighted based on the geometric proximity between the nodes, reinforces intra-community links and weakens inter-community links, making the clusters of the initial network more visible and more easily detectable. The geometric separation between the communities can become so strong that even a very simple parsing of the links may recover the communities as isolated components with surprisingly high precision. Furthermore, when used as a pre-processing step, our embedding and reweighting procedure can improve the performance of traditional community detection algorithms. Copyright © 2024, The Authors. All rights reserved.

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

Deep Reinforcement Learning-based Obstacle Avoidance for Robot Movement in Warehouse Environments

arXiv

引用

arXiv 2024年

作者： Li, Keqin Chen, Jiajing Yu, Denzhi Dajun, Tao Qiu, Xinyu Jieting, Lian Baiwei, Sun Shengyuan, Zhang Wan, Zhenyu Ji, Ran Hong, Bo Ni, Fanghao Department of Computer Science AMA University Quezon Philippines Courant Institute New York University 251 Mercer St New YorkNY10012 United States School of Information University of California Berkeley University Avenue and Oxford St BerkeleyCA94720 United States School of Engineering Carnegie Mellon University 5000 Forbes Avenue PittsburghPA United States College of Engineering Northeastern University Seattle United States New York University School of Professional Studies 50 West 4th Street New YorkNY United States School of Information and Computer Sciences University of California Irvine 510 Aldrich Hall IrvineCA United States College of Computing and Information Science Cornell University 616 Thurston Ave IthacaNY United States 2513 Las Palmas Ln PlanoTX75075 United States College of Engineering Cornell University 616 Thurston Ave IthacaNY14853 United States School of Informatics Computing and Cyber Systems Northern Arizona University FlagstaffAZ United States

At present, in most warehouse environments, the accumulation of goods is complex, and the management personnel in the control of goods at the same time with the warehouse mobile robot trajectory interaction, the traditional mobile robot can not be very good on the goods and pedestrians to feed back the correct obstacle avoidance strategy, in order to control the mobile robot in the warehouse environment efficiently and friendly to complete the obstacle avoidance task, this paper proposes a deep reinforcement learning based on the warehouse environment, the mobile robot obstacle avoidance Algorithm. Firstly, for the insufficient learning ability of the value function network in the deep reinforcement learning algorithm, the value function network is improved based on the pedestrian interaction, the interaction information between pedestrians is extracted through the pedestrian angle grid, and the temporal features of individual pedestrians are extracted through the attention mechanism, so that we can learn to obtain the relative importance of the current state and the historical trajectory state as well as the joint impact on the robot's obstacle avoidance strategy, which provides an opportunity for the learning of multi-layer perceptual machines afterwards. Secondly, the reward function of reinforcement learning is designed based on the spatial behaviour of pedestrians, and the robot is punished for the state where the angle changes too much, so as to achieve the requirement of comfortable obstacle avoidance;Finally, the feasibility and effectiveness of the deep reinforcement learning-based mobile robot obstacle avoidance algorithm in the warehouse environment in the complex environment of the warehouse are verified through simulation experiments. © 2024, CC BY.

关键词： Deep reinforcement learning

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：