|
|
|
题名
|
作者
|
年代
|
出处
|
被引量
|
| 1 | Boundary-processing-technique in EMD method and Hilbert transform显示文摘By virtue of neural network, a series of signals is extended forward and backward, as a result, two additional maxima and two additional minima are obtained at both ends of the original data set, with which the EMD decomposition can be exactly achieved with cubic spline interpolation. Meanwhile, by using of neural network every IMF component can also be extended forward and backward, which effectively restrains the end effect, thus the veracious Hilbert spectra are achieved. Verifications of the sample signals and the actual surface elevation of sea waves show that the present extension method is relatively accurate. | DENG Yongjun, WANG Wei, QIAN Chengchun, WANG Zhong & DAI DejunNetwork Center, Ocean University of Qingdao. Qingdao 266003, China Physical Oceanography Laboratory, Ocean Universily of Qingdao, Qingdao 266003, China | 2001 | Chinese Science Bulletin2001，46，11： | 91 |
| 2 | Convolutional neural networks for time series classification显示文摘Time series classification is an important task in time series data mining, and has attracted great interests and tremendous efforts during last decades. However, it remains a challenging problem due to the nature of time series data: high dimensionality,large in data size and updating continuously. The deep learning techniques are explored to improve the performance of traditional feature-based approaches. Specifically, a novel convolutional neural network(CNN) framework is proposed for time series classification. Different from other feature-based classification approaches,CNN can discover and extract the suitable internal structure to generate deep features of the input time series automatically by using convolution and pooling operations. Two groups of experiments are conducted on simulated data sets and eight groups of experiments are conducted on real-world data sets from different application domains. The final experimental results show that the proposed method outperforms state-of-the-art methods for time series classification in terms of the classification accuracy and noise tolerance. | Bendong Zhao Huanzhang Lu Shangfeng Chen Junliang Liu Dongya Wu | 2017 | Journal of Systems Engineering and Electronics2017，28，1： | 33 |
| 3 | Image segmentation of embryonic plant cell using pulse-coupled neural networks显示文摘Traditional image segmentation algorithms exhibit weak performance for plant cells which have complex structure. On the other hand, pulse-coupled neural network (PCNN) based on Eckhorn’s model of the cat visual cortex should be suitable to the segmentation of plant cell image. But the present theories cannot explain the relationship between the parameters of PCNN mathematical model and the effect of segmentation. Satisfactory results usually require time-consuming selection of experimental parameters. Mean-while, in a proper, selected parametric model, the number of iteration determines the segmented effect evaluated by visual judgment, which decreases the efficiency of image segmentation. To avoid these flaws, this note proposes a new PCNN algorithm for automatically segmenting plant embryonic cell image based on the maximum entropy principle. The algorithm produces a desirable result. In addition, a model with proper parameters can automatically determine the number of iteration, avoid visual judgment, | MA Yide DAI Rolan LI Lian WEI Lin | 2002 | Chinese Science Bulletin2002，47，2： | 28 |
| 4 | Machine learning methods for rockburst prediction-state-of-the-art review显示文摘One of the most serious mining disasters in underground mines is rockburst phenomena.They can lead to injuries and even fatalities as well as damage to underground openings and mining equipment.This has forced many researchers to investigate alternative methods to predict the potential for rockburst occurrence.However,due to the highly complex relation between geological,mechanical and geometric parameters of the mining environment,the traditional mechanics-based prediction methods do not always yield precise results.With the emergence of machine learning methods,a breakthrough in the prediction of rockburst occurrence has become possible in recent years.This paper presents a state-ofthe-art review of various applications of machine learning methods for the prediction of rockburst potential.First,existing rockburst prediction methods are introduced,and the limitations of such methods are highlighted.A brief overview of typical machine learning methods and their main features as predictive tools is then presented.The current applications of machine learning models in rockburst prediction are surveyed,with related mechanisms,technical details and performance analysis. | Yuanyuan Pu Derek B.Apel Victor Liu Hani Mitri | 2019 | International Journal of Mining Science and Technology2019，29，4： | 19 |
| 5 | Photogrammetry and Deep Learning显示文摘Deep learning has become popular and the mainstream technology in many researches related to learning,and has shown its impact on photogrammetry.According to the definition of photogrammetry,that is,a subject that researches shapes,locations,sizes,characteristics and inter-relationships of real objects from optical images,photogrammetry considers two aspects,geometry and semantics.From the two aspects,we review the history of deep learning and discuss its current applications on photogrammetry,and forecast the future development of photogrammetry.In geometry,the deep convolutional neural network(CNN)has been widely applied in stereo matching,SLAM and 3D reconstruction,and has made some effects but needs more improvement.In semantics,conventional methods that have to design empirical and handcrafted features have failed to extract the semantic information accurately and failed to produce types of“semantic thematic map”as 4D productions(DEM,DOM,DLG,DRG)of photogrammetry.This causes the semantic part of photogrammetry be ignored for a long time.The powerful generalization capacity,ability to fit any functions and stability under types of situations of deep leaning is making the automatic production of thematic maps possible.We review the achievements that have been obtained in road network extraction,building detection and crop classification,etc.,and forecast that producing high-accuracy semantic thematic maps directly from optical images will become reality and these maps will become a type of standard products of photogrammetry.At last,we introduce our two current researches related to geometry and semantics respectively.One is stereo matching of aerial images based on deep learning and transfer learning;the other is precise crop classification from satellite spatio-temporal images based on 3D CNN. | Jianya GONG Shunping JI | 2018 | Journal of Geodesy and Geoinformation Science2018，1，1： | 19 |
| 6 | STUDY ON MIXED MODEL OF NEURAL NETWORK FOR FARMLAND FLOOD/DROUGHT PREDICTION显示文摘The paper concerns a flood/drought prediction model involving the continuation of time seriesof a predictand and the physical factors influencing the change of predictand.Attempt is made toconstruct the model by the neural network scheme for the nonlinear mapping relation based onmulti-input and single output.The model is found of steadily higher predictive accuracy by testingthe output from one and multiple stepwise predictions against observations and comparing theresults to those from a traditional statistical model. | 金龙 罗莹 郭光 林振山 | 1997 | Acta meteorologica Sinica1997，11，3： | 18 |
| 7 | THE INFLUENCE OF CAVITY DAMAGE ON MECHANICAL PROPERTIES OF SUPERPLASTICALLY-DEFORMED MATERIALS AND ITS PREDICTION显示文摘Aluminium alloy LY12CZ sheet without any pre-treatment has been used to study the influence of cavity damage on the mechanical properties of the superplastically-deformed materials room temper- ature. The experimental results show that: (1) the lower the rate of cavitation as the superplastic strain increases, the higher the superplasticity is; (2) the mechanical properties of superplastically- deformed materials at room temperature decrease as the superplastically-deformed strain increases, especially, a noteworthy decrease in elasticity modulus, yield strenth and ultimate strength at room temperature begins to appear as the level of cavitation by area comes up to about 4%, while at the same level the reduction in area drops down a very large quantity. Then, a three-layer back- propagation neural network has been developed to predict the machanichl properties of the superplasti- cally-deformed material.The results acquired from the neural network are very | M. B. Liu, Z. C. Zhao,H. Gao,M. Q. Li and S. C. Wu College of Materials Science and Engineering, Northwestern Polytechnical University, Xi’an 710072, China | 2000 | Acta Metallurgica Sinica(English Letters)2000，13，2： | 18 |
| 8 | Advanced technology of high-resolution radar: target detection, tracking, imaging, and recognition显示文摘In recent years, the performances of radar resolution, coverage, and detection accuracy have been significantly improved through the use of ultra-wideband, synthetic aperture and digital signal processing technologies. High-resolution radars(HRRs) utilize wideband signals and synthetic apertures to enhance the range and angular resolutions of tracking, respectively. They also generate one-, two-, and even threedimensional high-resolution images containing the feature information of targets, from which the targets can be precisely classified and identified. Advanced signal processing algorithms in HRRs obtain important information such as range-Doppler imaging, phase-derived ranging, and micro-motion features. However, the advantages and applications of HRRs are restricted by factors such as the reduced signal-to-noise ratio(SNR)of multi-scatter point targets, decreased tracking accuracy of multi-scatter point targets, high demands of motion compensation, and low sensitivity of the target attitude. Focusing on these problems, this paper systematically introduces the novel technologies of HRRs and discusses the issues and solutions relevant to detection, tracking, imaging, and recognition. Finally, it reviews the latest progress and representative results of HRR-based research, and suggests the future development of HRRs. | Teng LONG Zhennan LIANG Quanhua LIU | 2019 | Science China(Information Sciences)2019，62，4： | 17 |
| 9 | Deep Learning Based Single Image Super-resolution:A Survey显示文摘Single image super-resolution has attracted increasing attention and has a wide range of applications in satellite imaging, medical imaging, computer vision, security surveillance imaging, remote sensing, objection detection, and recognition. Recently, deep learning techniques have emerged and blossomed, producing ' the state-of-the-art” in many domains. Due to their capability in feature extraction and mapping, it is very helpful to predict high-frequency details lost in low-resolution images. In this paper, we give an overview of recent advances in deep learning-based models and methods that have been applied to single image super-resolution tasks. We also summarize, compare and discuss various models from the past and present for comprehensive understanding and finally provide open problems and possible directions for future research. | Viet Khanh Ha Jin-Chang Ren Xin-Ying Xu Sophia Zhao Gang Xie Valentin Masero Amir Hussain | 2019 | International Journal of Automation and computing2019，16，4： | 17 |
| 10 | Satellite Image Matching Method Based on Deep Convolutional Neural Network显示文摘This article focuses on the first aspect of the album of deep learning: the deep convolutional method. The traditional matching point extraction algorithm typically uses manually designed feature descriptors and the shortest distance between them to match as the matching criterion. The matching result can easily fall into a local extreme value, which causes missing of the partial matching point. Targeting this problem, we introduce a two-channel deep convolutional neural network based on spatial scale convolution, which performs matching pattern learning between images to realize satellite image matching based on a deep convolutional neural network. The experimental results show that the method can extract the richer matching points in the case of heterogeneous, multi-temporal and multi-resolution satellite images, compared with the traditional matching method. In addition, the accuracy of the final matching results can be maintained at above 90%. | Dazhao FAN Yang DONG Yongsheng ZHANG | 2019 | Journal of Geodesy and Geoinformation Science2019，2，2： | 16 |
| 11 | A Short Text Classification Method Based on N-Gram and CNN显示文摘Text classification is a fundamental task in Nature language process(NLP) application. Most existing research work relied on either explicate or implicit text representation to settle this kind of problems, while these techniques work well for sentence and can not simply apply to short text because of its shortness and sparseness feature. Given these facts that obtaining the simple word vector feature and ignoring the important feature by utilizing the traditional multi-size filter Convolution neural network(CNN) during the course of text classification task, we offer a kind of short text classification model by CNN, which can obtain the abundant text feature by adopting none linear sliding method and N-gram language model, and picks out the key features by using the concentration mechanism, in addition employing the pooling operation can preserve the text features at the most certain as far as possible. The experiment shows that this method we offered, comparing the traditional machine learning algorithm and convolutional neural network, can markedly improve the classification result during the short text classification. | WANG Haitao HE Jie ZHANG Xiaohong LIU Shufen | 2020 | Chinese Journal of Electronics2020，29，2： | 15 |
| 12 | Application of deep learning method to Reynolds stress models of channel flow based on reduced-order modeling of DNS data显示文摘Recently,the methodology of deep learning is used to improve the calculation accuracy of the Reynolds-averaged Navier-Stokes (RANS) model.In this paper,a neural network is designed to predict the Reynolds stress of a channel flow of different Reynolds numbers.The rationality and the high efficiency of the neural network is validated by comparing with the results of the direct numerical simulation (DNS),the large eddy simulation (LES),and the deep neural network (DNN) of other studies.To further enhance the prediction accuracy,three methods are developed by using several algorithms and simplified models in the neural network.In the method 1.the regularization is introduced and it is found that the oscillation and the overfitting of the results are eflectively prevented.In the method 2,y^+ is embedded in the input variable while the combination of the invariants is simplified in the method 3.From the predicted results,it can be seen that by using the first two methods,the errors are reduced.Moreover,the method 3 shows considerable advantages in the DNS trend and the smoothness of a curve.Consequently,it is concluded that the DNNs can predict effectively the anisotropic Reynolds stress and is a promising technique of the computational fluid dynamics. | Zhen Zhang Xu-dong Song Shu-ran Ye Yi-wei Wang Chen-guang Huang Yi-ran An Yao-song Chen | 2019 | Journal of Hydrodynamics2019，31，1： | 13 |
| 13 | Composite Adaptive Control of Belt Polishing Force for Aero-engine Blade显示文摘The existing methods for blade polishing mainly focus on robot polishing and manual grinding.Due to the difficulty in high-precision control of the polishing force,the blade surface precision is very low in robot polishing,in particular,quality of the inlet and exhaust edges can not satisfy the processing requirements.Manual grinding has low efficiency,high labor intensity and unstable processing quality,moreover,the polished surface is vulnerable to burn,and the surface precision and integrity are difficult to ensure.In order to further improve the profile accuracy and surface quality,a pneumatic flexible polishing force-exerting mechanism is designed and a dual-mode switching composite adaptive control(DSCAC) strategy is proposed,which combines Bang-Bang control and model reference adaptive control based on fuzzy neural network(MRACFNN) together.By the mode decision-making mechanism,Bang-Bang control is used to track the control command signal quickly when the actual polishing force is far away from the target value,and MRACFNN is utilized in smaller error ranges to improve the system robustness and control precision.Based on the mathematical model of the force-exerting mechanism,simulation analysis is implemented on DSCAC.Simulation results show that the output polishing force can better track the given signal.Finally,the blade polishing experiments are carried out on the designed polishing equipment.Experimental results show that DSCAC can effectively mitigate the influence of gas compressibility,valve dead-time effect,valve nonlinear flow,cylinder friction,measurement noise and other interference on the control precision of polishing force,which has high control precision,strong robustness,strong anti-interference ability and other advantages compared with MRACFNN.The proposed research achieves high-precision control of the polishing force,effectively improves the blade machining precision and surface consistency,and significantly reduces the surface roughness. | ZHsAO Pengbing SHI Yaoyao | 2013 | Chinese Journal of Mechanical Engineering2013，26，5： | 12 |
| 14 | A Simple Hybrid Recursive Learning Algorithm with High Generalization Performance for Radial Basis Function Neural Network | ZHU Tao,\ WANG Zheng\|ou Institute of Systems Engineering, Tianjin University, Tianjin 300072, China | 2000 | Systems Science and Systems Engineering2000，10，1： | 12 |
| 15 | Identification of microRNAs and messenger RNAs involved in human umbilical cord mesenchymal stem cell treatment of ischemic cerebral infarction using integrated bioinformatics analysis显示文摘In recent years,a large number of differentially expressed genes have been identified in human umbilical cord mesenchymal stem cell(hUMSC)transplants for the treatment of ischemic cerebral infarction.These genes are involved in various biochemical processes,but the role of microRNAs(miRNAs)in this process is still unclear.From the Gene Expression Omnibus(GEO)database,we downloaded two microarray datasets for GSE78731(messenger RNA(mRNA)profile)and GSE97532(miRNA profile).The differentially expressed genes screened were compared between the hUMSC group and the middle cerebral artery occlusion group.Gene ontology enrichment and pathway enrichment analyses were subsequently conducted using the online Database for Annotation,Visualization,and Integrated Discovery.Identified genes were applied to perform weighted gene co-suppression analyses,to establish a weighted co-expression network model.Furthermore,the protein-protein interaction network for differentially expressed genes from turquoise modules was built using Cytoscape(version 3.40)and the most highly correlated subnetwork was extracted from the protein-protein interaction network using the MCODE plugin.The predicted target genes for differentially expressed miRNAs were also identified using the online database starBase v3.0.A total of 3698 differentially expressed genes were identified.Gene ontology analysis demonstrated that differentially expressed genes that are related to hUMSC treatment of ischemic cerebral infarction are involved in endocytosis and inflammatory responses.We identified 12 differentially expressed miRNAs in middle cerebral artery occlusion rats after hUMSC treatment,and these differentially expressed miRNAs were mainly involved in signaling in inflammatory pathways,such as in the regulation of neutrophil migration.In conclusion,we have identified a number of differentially expressed genes and differentially expressed mRNAs,miRNA-mRNAs,and signaling pathways involved in the hUMSC treatment of ischemic cerebral infarction.Bioinformatics and interaction analyses can provide novel clues for further research into hUMSC treatment of ischemic cerebral infarction. | Yin-Meng Qu Xin Sun Xiu-Li Yan Hang Jin Zhen-Ni Guo Yi Yang | 2019 | Neural Regeneration Research2019，14，9： | 12 |
| 16 | Forecasting Different Types of Convective Weather: A Deep Learning Approach显示文摘A deep learning objective forecasting solution for severe convective weather(SCW) including short-duration heavy rain(HR), hail, convective gusts(CG), and thunderstorms based on numerical weather prediction(NWP) data was developed. We first established the training datasets as follows. Five years of severe weather observations were utilized to label the NCEP final(FNL) analysis data. A large number of labeled samples for each type of weather were then selected for model training. The local temperature, pressure, humidity, and winds from 1000 to 200 h Pa, as well as dozens of convective physical parameters, were taken as predictors in our model. A six-layer convolutional neural network(CNN) model was then built and trained to obtain optimal model weights. After that, the trained model was used to predict SCW based on the Global Forecast System(GFS) forecast data as input. The performances of the CNN model and other traditional methods were compared. The results show that the deep learning algorithm had a higher classification accuracy on HR and hail than support vector machine, random forests, and other traditional machine learning algorithms. The objective forecasts by use of the deep learning algorithm also showed better forecasting skills than the subjective forecasts by the forecasters. The threat scores(TSs) of thunderstorm, HR, hail, and CG were increased by 16.1%, 33.2%, 178%, and 55.7%, respectively. The deep learning forecast model is currently used in the National Meteorological Center of China to provide guidance for the operational SCW forecasting over China. | Kanghui ZHOU Yongguang ZHENG Bo LI Wansheng DONG Xiaoling ZHANG | 2019 | Journal of Meteorological Research2019，33，5： | 12 |
| 17 | Online scheduling of image satellites based on neural networks and deep reinforcement learning显示文摘In the ‘‘Internet Plus' era, space-based information services require effective and fast image satellite scheduling. Most existing studies consider image satellite scheduling to be an optimization problem to solve with searching algorithms in a batch-wise manner. No real-time speed method for satellite scheduling exists. In this paper, with the idea of building a real-time speed method, satellite scheduling is remodeled based on a Dynamic and Stochastic Knapsack Problem(DSKP), and the objective is to maximize the total expected profit. No existing algorithm could be able to solve this novel scheduling problem properly. With inspiration from the recent achievements in Deep Reinforcement Learning(DRL) in video games, AlphaGo and dynamic controlling,a novel DRL-based method is applied to training a neural network to schedule tasks. The numerical results show that the method proposed in this paper can achieve relatively good performance with real-time speed and immediate respond style. | Haijiao WANG Zhen YANG Wugen ZHOU Dalin LI | 2019 | Chinese Journal of Aeronautics2019，32，4： | 12 |
| 18 | Application of artificial intelligence in gastroenterology显示文摘Artificial intelligence(AI) using deep-learning(DL) has emerged as a breakthrough computer technology. By the era of big data, the accumulation of an enormous number of digital images and medical records drove the need for the utilization of AI to efficiently deal with these data, which have become fundamental resources for a machine to learn by itself. Among several DL models, the convolutional neural network showed outstanding performance in image analysis. In the field of gastroenterology, physicians handle large amounts of clinical data and various kinds of image devices such as endoscopy and ultrasound. AI has been applied in gastroenterology in terms of diagnosis,prognosis, and image analysis. However, potential inherent selection bias cannot be excluded in the form of retrospective study. Because overfitting and spectrum bias(class imbalance) have the possibility of overestimating the accuracy,external validation using unused datasets for model development, collected in a way that minimizes the spectrum bias, is mandatory. For robust verification,prospective studies with adequate inclusion/exclusion criteria, which represent the target populations, are needed. DL has its own lack of interpretability.Because interpretability is important in that it can provide safety measures, help to detect bias, and create social acceptance, further investigations should be performed. | Young Joo Yang Chang Seok Bang | 2019 | World Journal of Gastroenterology2019，25，14： | 12 |
| 19 | Forest Fire Susceptibility Modeling Using a Convolutional Neural Network for Yunnan Province of China显示文摘Forest fires have caused considerable losses to ecologies, societies, and economies worldwide. To minimize these losses and reduce forest fires, modeling and predicting the occurrence of forest fires are meaningful because they can support forest fire prevention and management. In recent years, the convolutional neural network(CNN) has become an important state-of-the-art deep learning algorithm, and its implementation has enriched many fields. Therefore, we proposed a spatial prediction model for forest fire susceptibility using a CNN. Past forest fire locations in Yunnan Province, China, from 2002 to 2010, and a set of 14 forest fire influencing factors were mapped using a geographic information system. Oversampling was applied to eliminate the class imbalance, and proportional stratified sampling was used to construct the training/validation sample libraries. A CNN architecture that is suitable for the prediction of forest fire susceptibility was designed and hyperparameters were optimized to improve the prediction accuracy. Then, the test dataset was fed into the trained model to construct the spatial prediction map of forest fire susceptibility in Yunnan Province.Finally, the prediction performance of the proposed model was assessed using several statistical measures—Wilcoxon signed-rank test, receiver operating characteristic curve,and area under the curve(AUC). The results confirmed the higher accuracy of the proposed CNN model(AUC 0.86)than those of the random forests, support vector machine,multilayer perceptron neural network, and kernel logistic regression benchmark classifiers. The CNN has stronger fitting and classification abilities and can make full use of neighborhood information, which is a promising alternative for the spatial prediction of forest fire susceptibility. This research extends the application of CNN to the prediction of forest fire susceptibility. | Guoli Zhang Ming Wang Kai Liu | 2019 | International Journal of Disaster Risk Science2019，10，3： | 12 |
| 20 | Study on Ann-Based Multi-Step Prediction Model of Short-Term Climatic Variation显示文摘In the context of 1905-1995 series from Nanjing and Hangzhou, study is undertaken of establishing a predictive model of annual mean temperature in 1996-2005 to come over the Changjiang (Yangtze River) delta region through mean generating function and artificial neural network in combination. Results show that the established model yields mean error of 0.45℃ for their absolute values of annual mean temperature from 10 yearly independent samples (1986-1995) and the difference between the mean predictions and related measurements is 0. 156℃, The developed model is found superior to a mean generating function regression model both in historical data fitting and independent sample prediction. | 金龙 居为民 缪启龙 | 2000 | Advances in Atmospheric Sciences2000，17，1： | 11 |
您所在的机构暂未开通此部分的功能，请联系管理员