检索结果-内蒙古大学图书馆

International Joint Conference on Neural Networks

作者： Can Wang Qiudan Li Lei Wang Daniel Dajun Zeng The State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Department of Management Information Systems University of Arizona

ISBN: (纸本)9781509061839

With the rapid growth of Web 2.0, social media has become a prevalent information sharing and spreading platform, where users can retweet interesting messages. To better understand the propagation mechanism for information diffusion, it is necessary to model the user retweeting behavior and predict future retweets. Some existing work in retweeting prediction based on matrix factorization focuses on using user-message interaction information, user information and social influence information, etc. The challenge of improving prediction performance is how to jointly perform deep representation of these information to solve the sparsity problem and then learn a more comprehensive retweeting behavior model. Inspired by word2vec and co-factor matrix factorization model, this paper proposes a hybrid model, called HCFMF, for learning users' retweeting behavior, it first computes the message content similarity by considering the message co-occurrence, the author information and word2vec based low-dimensional representation of content, then, jointly decomposes the user-message matrix and message-message similarity matrix based on a co-factorization model. We empirically evaluate the performance of the proposed model on real world weibo datasets. Experimental results show that taking the dense representation of author and content information into consideration could allow us make more accurate analysis of users' retweeting patterns. The mined patterns could serve as a feedback channel for both consumers and management departments.

关键词： Retweeting prediction Co-factor matrix factorization Word2vec Low-dimensional representation matrix decomposition Consumers feedback channel Hybrid modes information role model Information Sharing

来源：评论

学校读者我要写书评

暂无评论

Relative Torque Contribution Based Model Simplification for Robotic Dynamics Identification

Relative Torque Contribution Based Model Simplification for ...

引用

IEEE Symposium Series on Computational Intelligence

作者： Weiqun Wang Zeng-Guang Hou Xu Liang Shixin Ren Liang Peng Lincong Luo Chengkun Cui State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China

It has been proved that minimizing the condition number of the observation matrix, which is calculated from the robot dynamic model and the associated exciting trajectories, is very effective for improving the identification accuracy of robotic dynamics. A relative simple dynamic model is beneficial for reduction of the associated condition number, and hence, several model simplification methods have been proposed in the literature. However, the existed methods cannot be used to efficiently process model structural errors, which will inevitably cause inaccurate estimation of the dynamics. Therefore, a novel model simplification method based on relative contribution of the undetermined parameters, is proposed to overcome the deficiency. Firstly, exciting trajectories for model simplification are designed by using finite Fourier series and optimized by using the condition number criteria. Then, the optimized exciting trajectory is implemented on the robot, and joint torques and motion data are recorded, which are used to calculate relative contribution of the undetermined parameters to joint torques. The model can be simplified repeatedly by neglecting the parameter that contributes least until the condition number is small enough. Finally, the performance of the proposed method is demonstrated by the identification and validation experiments conducted on a lower limb rehabilitation robot.

关键词： Identification Model Simplification Robotic Dynamics

来源：评论

学校读者我要写书评

暂无评论

Robust kernelized multi-view self-representations for clustering by tensor multi-rank minimization

arXiv

引用

arXiv 2017年

作者： Qu, Yanyun Liu, Jinyan Xie, Yuan Zhang, Wensheng Fujian Key Laboratory of Sensing and Computing for Smart City School of Information Science and Engineering Xiamen University China State Key Laboratory of Intelligent Control Management of Complex Systems Chinese Academy of Sciences

Most recently, tensor-SVD is implemented on multi-view self-representation clustering and has achieved the promising results in many real-world applications such as face clustering, scene clustering and generic object clustering. However, tensor-SVD based multi-view self-representation clustering is proposed originally to solve the clustering problem in the multiple linear subspaces, leading to unsatisfactory results when dealing with the case of non-linear subspaces. To handle data clustering from the non-linear subspaces, a kernelization method is designed by mapping the data from the original input space to a new feature space in which the transformed data can be clustered by a multiple linear clustering method. In this paper, we make an optimization model for the kernelized multi-view self-representation clustering problem. We also develop a new efficient algorithm based on the alternation direction method and infer a closed-form solution. Since all the subproblems can be solved exactly, the proposed optimization algorithm is guaranteed to obtain the optimal solution. In particular, the original tensor-based multi-view self-representation clustering problem is a special case of our approach and can be solved by our algorithm. Experimental results on several popular real-world clustering datasets demonstrate that our approach achieves the state-of-The-Art performance. Copyright © 2017, The Authors. All rights reserved.

关键词： Tensors

来源：评论

学校读者我要写书评

暂无评论

Decentralized adaptive neural network sliding mode position/force control of constrained reconfigurable manipulators

引用

Journal of Central South University 2016年第11期23卷 2917-2925页

作者：李元春丁贵彬赵博 Department of Control Science and Engineering Changchun University of Technology Department of Control Science and Engineering Jilin University State Key Laboratory of Management and Control for Complex Systems Institute of AutomationChinese Academy of Sciences

A decentralized adaptive neural network sliding mode position/force control scheme is proposed for constrained reconfigurable manipulators. Different from the decentralized control strategy in multi-manipulator cooperation, the proposed decentralized position/force control scheme can be applied to series constrained reconfigurable manipulators. By multiplying each row of Jacobian matrix in the dynamics by contact force vector, the converted joint torque is obtained. Furthermore, using desired information of other joints instead of their actual values, the dynamics can be represented as a set of interconnected subsystems by model decomposition technique. An adaptive neural network controller is introduced to approximate the unknown dynamics of subsystem. The interconnection and the whole error term are removed by employing an adaptive sliding mode term. And then, the Lyapunov stability theory guarantees the stability of the closed-loop system. Finally, two reconfigurable manipulators with different configurations are employed to show the effectiveness of the proposed decentralized position/force control scheme.

关键词： constrained reconfigurable manipulators position/force control model decomposition decentralized control neural network

来源：评论

学校读者我要写书评

暂无评论

Traffic Flow Data Forecasting Based on Interval Type-2 Fuzzy Sets Theory

引用

IEEE/CAA Journal of Automatica Sinica 2016年第2期3卷 141-148页

作者： Runmei Li Chaoyang Jiang Fenghua Zhu Xiaolong Chen Beijing Jiaotong University the State Key Laboratory for Management and Control of Complex Systems Institute of Automation Chinese Academy of Sciences

This paper proposes a long-term forecasting scheme and implementation method based on the interval type-2 fuzzy sets theory for traffic flow data. The type-2 fuzzy sets have advantages in modeling uncertainties because their membership functions are fuzzy. The scheme includes traffic flow data preprocessing module, type-2 fuzzification operation module and long-term traffic flow data forecasting output module, in which the Interval Approach acts as the core algorithm. The central limit theorem is adopted to convert point data of mass traffic flow in some time range into interval data of the same time range (also called confidence interval data) which is being used as the input of interval approach. The confidence interval data retain the uncertainty and randomness of traffic flow, meanwhile reduce the influence of noise from the detection data. The proposed scheme gets not only the traffic flow forecasting result but also can show the possible range of traffic flow variation with high precision using upper and lower limit forecasting result. The effectiveness of the proposed scheme is verified using the actual sample application. © 2014 Chinese Association of Automation.

关键词： Data handling Forecasting Fuzzy sets Membership functions Uncertainty analysis

来源：评论

学校读者我要写书评

暂无评论

Motion characterization of a relaxor-based single crystals actuator for Na-nomanipulation system

Motion characterization of a relaxor-based single crystals a...

引用

Youth Academic Annual Conference of Chinese Association of Automation (YAC)

作者： Tianlu Zhang Zhangming Du Chao Zhou Lu Deng Zhiqiang Cao Shuo Wang Min Tan State Key Laboratory of Management and Control for Complex Systems University of Chinese Academy of Sciences Beijing China State Key Laboratory of Management and Control for Complex Systems Chinese Academy of Sciences Beijing China School of Statistics and Mathematics Central University of Finance and Economics Beijing China

ISBN: (纸本)9781538629024

The micro-operation technology is developing rapidly since the miniaturization of the research object and the precision of the operation. As the actuator of micro-operation, the piezoelectric material with the inverse piezoelectric effect has been widely used in this field. A new type of actuator based on piezoelectric single crystals is characterized in this paper, and a set of test system is designed to obtain the displacement of the actuator under different voltage waveforms. The relationship between the induced voltage and output displacement is analyzed. It is found that the dynamic voltage and the output displacement show a good linear relationship, and the hysteresis of the motion is smaller than the traditional piezoelectric ceramics. This good performance of the piezoelectric single crystal actuator is expected to directly use open-loop control, greatly improve the actuator control frequency and accuracy, which can reduce the complexity of the application system.

关键词： Crystals Actuators Ceramics Hysteresis Sensors Electric fields Drives

来源：评论

学校读者我要写书评

暂无评论

Spike-burst chimera states in an adaptive exponential integrate-and-fire neuronal network

arXiv

引用

arXiv 2019年

作者： Santos, Moises S. Protachevicz, Paulo R. Iarosz, Kelly C. Caldas, Iberê L. Viana, Ricardo L. Borges, Fernando S. Ren, Hai-Peng Szezech, José D. Batista, Antonio M. Grebogi, Celso Department of Physics Federal University of Paraná CuritibaPR80060-000 Brazil Graduate in Science Program - Physics State University of Ponta Grossa Ponta Grossa PR84030-900 Brazil Institute of Physics University of São Paulo São PauloSP05508-900 Brazil Center for Mathematics Computation and Cognition Federal University of ABC São Bernardo do CampoSP09606-045 Brazil Shaanxi Key Laboratory of Complex System Control and Intelligent Information Processing Xian University of Technology Xi'an710048 China Xian Technological University Xi'an710021 China Graduate in Science Program - Physics State University of Ponta Grossa Ponta Grossa PR84030-900 Brazil Institute for Complex Systems and Mathematical Biology King's College University of Aberdeen AberdeenAB24 3UE United Kingdom

Chimera states are spatiotemporal patterns in which coherence and incoherence coexist. We observe the coexistence of synchronous (coherent) and desynchronous (incoherent) domains in a neuronal network. The network is composed of coupled adaptive exponential integrate-and-fire neurons that are connected by means of chemical synapses. In our neuronal network, the chimera states exhibit spatial structures both with spikes and bursts activities. Furthermore, those desynchronised domains not only have either spike or burst activity, but we show that the structures switch between spikes and bursts as the time evolves. Moreover, we verify the existence of multicluster chimera states. Copyright © 2019, The Authors. All rights reserved.

关键词： Neural networks

来源：评论

学校读者我要写书评

暂无评论

End-to-End Driving Activities and Secondary Tasks Recognition Using Deep Convolutional Neural Network and Transfer Learning

End-to-End Driving Activities and Secondary Tasks Recognitio...

引用

2018 IEEE intelligent Vehicles Symposium, IV 2018

作者： Xing, Yang Tang, Jianlin Liu, Hong Lv, Chen Cao, Dongpu Velenis, Efstathios Wang, Fei-Yue Advanced Vehicle Engineering Centre Cranfield University BedfordMK43 0AL United Kingdom Vehicle Intelligence Pioneers Ltd Qingdao China Jiangsu XCMG Construction Machinery Research Institute Ltd. No. 26 Tuolanshan Road Economic Development Zone Xuzhou Jiangsu221004 China School of Mechanical and Mechatronics Engineering University of Waterloo 200 University Avenue West WaterlooONN2L 3G1 Canada State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 China

ISBN: (纸本)9781538644522

Drivers' decision and their corresponding behaviors are important aspects that can affect the driving safety, and it is necessary to understand the driver behaviors in real-time. In this study, an end-to-end driving-related tasks recognition system is proposed. Specifically, seven common driving activities are identified, which are normal driving, right mirror checking, rear mirror checking, left mirror checking, using in-vehicle video device, texting, and answering mobile phone. Among these, the first four activities are regarded as normal driving tasks, while the rest three are divided into distraction group. The images are collected using a consumer range camera, namely, Kinect. In total, five drivers are involved in the naturalistic data collection. Before training the identification model, the raw images are first segmented using a Gaussian mixture model (GMM) to extract the driver region from the background. Then, a pre-trained deep convolutional neural network (CNN) model is trained to classify the behaviors, which directly takes the processed RGB images as the input and outputs the identified label. In this work, the AIexNet is selected as the pre-trained CNN model. Then, to reduce the training cost, the transfer learning mechanism is applied to the CNN model. An average of 79\% detection accuracy is achieved for the seven driving tasks. The proposed integration model can be used as a low-cost driver distraction and dangerous tasks recognition modeL. © 2018 IEEE.

关键词： Neural network models

来源：评论

学校读者我要写书评

暂无评论

Inside Cover: A novel in vivo Cerenkov luminescence image-guided surgery on primary and metastatic colorectal cancer (J. Biophotonics 3/2020)

引用

Journal of Biophotonics 2020年第3期13卷

作者： Zeyu Zhang Yawei Qu Yu Cao Xiaojing Shi Hongbo Guo Xiaojun Zhang Sheng Zheng Haifeng Liu Zhenhua Hu Jie Tian Engineering Research Center of Molecular and Neuro Imaging of Ministry of Education School of Life Science and Technology Xidian University Xi'an China CAS Key Laboratory of Molecular Imaging Beijing Key Laboratory of Molecular Imaging The State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing China Department of Gastroenterology the Third Medical Centre Chinese PLA General Hospital Beijing China Department of Control Science and Engineering Harbin Institute of Technology Harbin China Department of Anorectal the Third medical Centre Chinese PLA General Hospital Beijing China University of Chinese Academy of Sciences Beijing China School of Information Sciences and Technology Northwest University Xi'an China Department of Nuclear Medicine Chinese PLA General Hospital Beijing China Beijing Advanced Innovation Center for Big Data-Based Precision Medicine School of Medicine Beihang University Beijing China

This study presents a novel intraoperative in vivo imaging approach which harnessed Cerenkov luminescence (CL) to detect primary and metastatic colorectal cancer (CRC) using clinically approved radiopharmaceuticals. In the mice and swine experiments, the proposed approach effectively improved the effect of CRC surgery. The approach is believed to be promising for utilizing CL in open surgery. Further details can be found in the article by Zeyu Zhang, Yawei Qu, Yu Cao et al. ( e201960152 )

关键词：

来源：评论

学校读者我要写书评

暂无评论

Dual refinement networks for accurate and fast object detection in real-world scenes

arXiv

引用

arXiv 2018年

作者： Chen, Xingyu Yu, Junzhi Kong, Shihan Wu, Zhengxing Wen, Li State Key Laboratory of Management and Control for Complex Systems Institute of Automation Chinese Academy of Sciences Beijing100190 University of Chinese Academy of Sciences Beijing100049 China Beijing Innovation Center for Engineering Science and Advanced Technology Peking University Beijing100871 China School of Mechanical Engineering and Automation Beihang University Beijing100191 China

Object detection has been vigorously studied for years but fast accurate detection for real-world scenes remains a very challenging problem. Overcoming drawbacks of singlestage detectors, we take aim at precisely detecting objects for static and temporal scenes in real time. Firstly, as a dual refinement mechanism, a novel anchor-offset detection including an anchor refinement, a feature offset refinement, and a deformable detection head is designed for two-step regression and capture of accurate detection features. Based on the anchoroffset detection, a dual refinement network (DRN) is developed for high-performance static detection, where a multi-deformable head is further designed to leverage contextual information for describing objects. As for temporal detection in real-world scenes, temporal refinement networks (TRN) and temporal dual refinement networks (TDRN) are developed by propagating the refinement information across time, where we also propose a loose refinement strategy to match object motion with the previous refinement. Our proposed methods are evaluated on PASCAL VOC, COCO, and ImageNet VID datasets. Extensive comparison on static and temporal detection verify the superiority of DRN, TRN, and TDRN. Consequently, our developed approaches run in a fairly fast speed, and in the meantime achieve a significantly enhanced detection accuracy, i.e., a mAP of 82.8% on VOC 2007, 80.6% on VOC 2012, 69.4% on VID 2017, and an AP of 34.3% on COCO. Ultimately, producing encouraging results, our methods are applied to underwater object grasping with an autonomous system. Codes are publicly available at http://***/SeanChenxy/TDRN. Copyright © 2018, The Authors. All rights reserved.

关键词： Computer vision

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：