检索结果-内蒙古大学图书馆

International Joint Conference on Neural Networks (IJCNN)

作者： Yousefi, Ali Dibazar, Alireza A. Berger, Theodore W. Univ So Calif Neural Dynam Lab Los Angeles CA 90089 USA

ISBN: (纸本)9781424496365

The main focus of this paper is to introduce a new supervised learning algorithm for spiking neural networks. The learning algorithm minimizes the overall differences between spike times of target and test spike trains by utilizing a new quantitative similarity measure which has been defined in this work. The actual membrane potential of a post-synaptic neuron is adjusted at the time of spikes based on what has been measured from similarity measure in order to generate the desired membrane potential. Finally, by utilizing gradient descent algorithm, the parameters of the spiking neural network are tuned to generate the desired output membrane potential. The proposed algorithm was applied to tune the facilitation, depression, and synaptic weight constants of the Dynamic Synapses Neural Network - DSNN - for the aim of input-output functional mapping. The simulation results show that the system identification task converges to the global optimum. The rate-to-time coding simulation performs with more than 75 percent accuracy. The performance of both system identification and rate-to-time coding is due to adaptation of short and long term synaptic parameters which cannot be accomplished if only synaptic weight is adapted.

关键词： Adaptation models Calcium Energy states Filtering Heuristic algorithms Mathematical model Neurons depression weight constant facilitation weight constant gradient descent algorithm gradient methods input-output functional mapping learning (artificial intelligence) neural nets post-synaptic neuron rate-to-time coding simulation single layer dynamic synapses neural network spike train supervised learning algorithm synaptic weight constant

来源：评论

学校读者我要写书评

暂无评论

Partial discharge pattern recognition using variable predictive model-based class discrimination with kernel partial least squares regression

引用

IET SCIENCE MEASUREMENT & TECHNOLOGY 2018年第3期12卷 360-367页

作者： Jia, Yafei Zhu, Yongli North China Elect Power Univ State Key Lab Alternate Elect Power Syst Renewabl Baoding 071003 Peoples R China

The pattern recognition of partial discharge (PD) is critical to evaluate the insulation condition and locate the defect of a power transformer. The existing pattern recognition methods fail to make use of the inter-relations of the extracted features of PD signals. In fact, the inter-relations can show distinct dissimilarities between different classes of the signals. To overcome the defect of existing pattern recognition methods, the variable predictive model-based class discrimination (VPMCD), a new pattern recognition method, is introduced for the pattern recognition of PD in this study. However, the original VPMCD lacks general expression ability of the inter-relations of extracted features and inherits the shortcomings of least squares (LS) regression. To overcome the above defects, an improved VPMCD based on kernel partial LS (KPLS) regression, which is called as KPLS-VPMCD, is proposed in this study. Experiments and analyses are implemented using both UCI datasets and the extracted features of PD signals. The experiments show that the performance of the proposed KPLS-VPMCD is better than those of the existing methods such as VPMCD, back propagation neural networks, and support vector machines. The conclusion is that KPLS-VPMCD is an efficient supervised learning algorithm with consistency and good performance for PD pattern recognition.

关键词： partial discharge measurement pattern recognition prediction theory least squares approximations regression analysis power transformer insulation feature extraction partial discharge pattern recognition method variable predictive model-based class discrimination PD pattern recognition method insulation condition evaluation power transformer feature extraction LS regression improved VPMCD kernel partial least square regression KPLS regression UCI dataset intelligent patter recognition method backpropagation neural network support vector machine supervised learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Planetary gearbox fault diagnostic method using acoustic emission sensors

引用

IET SCIENCE MEASUREMENT & TECHNOLOGY 2015年第8期9卷 936-944页

作者： Yoon, Jae He, David Univ Illinois Dept Mech & Ind Engn Chicago IL 60607 USA

In this study, a new acoustic emission (AE) sensor-based planetary gearbox (PGB) fault diagnosis method is presented. The method includes a heterodyne-based AE data acquisition system, empirical mode decomposition (EMD)-based AE signal analysis method, and computation of condition indicators (CIs) for PGB fault diagnosis. The heterodyne technique is hardware-implemented to downshift the sampling frequency of AE signals at a rate compatible to vibration analysis. The sampled AE signals are processed using EMD to extract PGB fault features and compute the CIs. The CIs are input into supervised learning algorithms for PGB fault diagnosis. The method is validated on a set of seeded localised faults on all gears: sun gear, planetary gear, and ring gear. The validation results have shown a promising PGB fault diagnostic performance using the presented method.

关键词： gears fault diagnosis acoustic emission acoustic signal processing singular value decomposition heterodyne detection acoustic signal detection data acquisition condition monitoring signal sampling vibrations feature extraction learning (artificial intelligence) electric sensing devices acoustic devices acoustic emission sensor planetary gearbox fault diagnosis method heterodyne-based AE data acquisition system empirical mode decomposition EMD-based AE signal analysis method condition indicator computation PGB fault diagnosis sampling frequency vibration analysis AE signal processing PGB fault feature extraction supervised learning algorithm seeded localised fault sun gear planetary gear ring gear

来源：评论

学校读者我要写书评

暂无评论

Adaptive nonlinear equalizer with reduced computational complexity

引用

SIGNAL PROCESSING 1995年第3期47卷 307-317页

作者： Oh, DG Choi, JY Lee, CW Satellite Broadcasting System Section Electronics and Telecommunications Research Institute P. O. Box 106 Taejon 305-306 South Korea School of Electrical Engineering Seoul National University Seoul South Korea

This paper presents an approach of adaptive nonlinear equalization having much less computational complexity than existing nonlinear equalizers. The proposed equalizer first estimates many local regions in the symbol decision space, and then performs a biased finite impulse response filtering in each estimated local region, The final estimate of the transmitted symbol is obtained by a weighted average of the local results, By using a self-organizing and localized linear tuning methods, the proposed nonlinear equalizer can achieve an almost optimal decision making capability like the existing nonlinear equalizers, Simulation results show that the proposed equalizer can be effectively applied to on-line adaptation environments due to its reduced computational complexity, making complexity similar to that of a linear equalizer.

关键词： adaptive nonlinear equalizers self-organizing method supervised learning algorithm weighted average biased finite impulse response filter

来源：评论

学校读者我要写书评

暂无评论

Car Driver's Sympathetic Reaction Detection Through Electrodermal Activity and Electrocardiogram Measurements

引用

IEEE TRANSACTIONS ON BIOMEDICAL ENGINEERING 2020年第12期67卷 3413-3424页

作者： Zontone, Pamela Affanni, Antonio Bernardini, Riccardo Piras, Alessandro Rinaldo, Roberto Formaggia, Fabio Minen, Diego Minen, Michela Savorgnan, Carlo Univ Udine Dipartimento Politecn Ingn & Architetura Udine Italy VI Grade Srl I-33010 Tavagnacco Italy

Objective: in this paper we propose a system to detect a subject's sympathetic reaction, which is related to unexpected or challenging events during a car drive. Methods: we use the Electrocardiogram (ECG) signal and the Skin Potential Response (SPR) signal, which has several advantages with respect to other Electrodermal (EDA) signals. We record one SPR signal for each hand, and use an algorithm that, selecting the smoother signal, is able to remove motion artifacts. We extract statistical features from the ECG and SPR signals in order to classify signal segments and identify the presence or absence of emotional events via a supervised learning algorithm. The experiments were carried out in a company which specializes in driving simulator equipment, using a motorized platform and a driving simulator. Different subjects were tested with this setup, with different challenging events happening on predetermined locations on the track. Results: we obtain an Accuracy as high as 79.10% for signal blocks and as high as 91.27% for events. Conclusion: results demonstrate the good performance of the presented system in detecting sympathetic reactions, and the effectiveness of the motion artifact removal procedure. Significance: our work demonstrates the possibility to classify the emotional state of the driver, using the ECG and EDA signals and a slightly invasive setup. In particular, the proposed use of SPR and of the motion artifact removal procedure are crucial for the effectiveness of the system.

关键词： Stress Electrocardiography Skin Motion artifacts Biomedical measurement Feature extraction Support vector machines Stress detection skin potential response motion artifact electrocardiogram supervised learning algorithm

来源：评论

学校读者我要写书评

暂无评论

Image receptive fields for artificial neural networks

引用

NEUROCOMPUTING 2014年 144卷 258-270页

作者： Buessler, Jean-Luc Smagghe, Philippe Urban, Jean-Philippe Univ Haute Alsace Fac Sci & Tech F-68093 Mulhouse France

This paper describes the structure of the Image Receptive Fields Neural Network (IRF-NN) introduced recently by our team. This structure extends simplified learning introduced by Extreme learning Machine and Reservoir Computing techniques to the field of images. Neurons are organized in a single hidden layer feedforward network architecture with an original organization of the network's input weights. To represent color images efficiently, without prior feature extraction, the weight values linked to a neuron are determined by a 2-D Gaussian function. The activation of a neuron by an image presents the properties of a nonlinear localized receptive field, parameterized with a small number of degrees of freedom. A network composed of a large number of neurons, each associated with a randomly initialized and constant receptive field, induces a remarkable representation of the images. supervised training determines only the output weights of the network. It is therefore extremely fast, without retro-propagation or iterations, adapted to large sets of images. The network is easy to implement, presents excellent generalization performances for classification applications, and allows the detection of unknown inputs. The efficiency of this technique is illustrated with several benchmarks, photo and video datasets. (C) 2014 Elsevier B.V. All rights reserved.

关键词： Computer vision Visual object recognition Artificial neural networks model supervised learning algorithm Visual receptive fields Extreme learning Machine

来源：评论

学校读者我要写书评

暂无评论

Building the carbonate pore-type classifier for well logging via the blended training dataset

引用

ACTA GEOPHYSICA 2021年第1期69卷 77-94页

作者： Li Bonan Si Wenpeng Shen Hui Zhang Xuebing Nanjing Tech Univ Coll Transportat Sci & Engn Nanjing 210009 Peoples R China Sinopec Geophys Res Inst Nanjing 211103 Peoples R China Jilin Jianzhu Univ Sch Geomat & Prospecting Engn Changchun 130118 Peoples R China

The exploration of carbonate rocks has outstanding economic benefits, as well as facing the extreme challenge of reservoir characterization. This article has proposed a data-based description scheme generalizing carbonate pore-type characteristics from both laboratory measurements and theoretical predictions to the well logging dataset. Firstly, in the feature space of elastic properties, we employed the supervised machine learning (ML) algorithm to convert this pore-type classification process from a typical nonlinear inversion to sample label allocation problem. Secondly, to alleviate the inherent scale gaps between data sources, virtual samples were randomly mixed into the laboratory measured dataset. Through inheriting or mimicking statistical elastic features of limited core samples, the new built training dataset could improve the overall sample richness and thus help the ML algorithms making better identification decisions. On the one hand, this scheme was verified by 74 carbonate samples. In the feature space of high dimensions, the blended dataset trained radial basis function support vector machine accurately separated different carbonate pore systems. Moreover, using logging curves of a carbonate gas field, we verified the generalization capability of this scheme over unbalanced data scales. Searching skills were used to optimize model and classifier setups according logging curves of a specific interval. Finally, with the help of the vertical label distributions, logging elastic response modes and historical pore evolution footprints were further studied.

关键词： Carbonate pore-type effect supervised learning algorithm Logging pore-type classification

来源：评论

学校读者我要写书评

暂无评论

Quality-Based Ranking of Translation Outputs

引用

IT PROFESSIONAL 2020年第4期22卷 21-27页

作者： Bharti, Nivedita Joshi, Nisheeth Mathur, Iti Katyayan, Pragya Banasthali Vidyapith Vanasthali India Banasthali Vidyapith Dept Comp Sci Vanasthali India

Translation ranking is inherently of great significance for machine translation (MT), as it allows the comparison of performances of multiple MT systems as well as for its efficient training. This article demonstrates a mechanism that is used for ranking the translation outputs generated by the MT systems from best to worst. To implement this approach, the system exploits a supervised learning algorithm trained over existing manual ranking by using various features obtained after the linguistic analysis of both source and target side sentences without relying on the reference translation.

关键词： Language Translation learning Artificial Intelligence Natural Language Processing Reference Translation supervised learning algorithm Multiple MT Systems Machine Translation Translation Ranking Quality Based Ranking Feature Extraction supervised learning Task Analysis Ranking Statistics Predictive Models Linguistics

来源：评论

学校读者我要写书评

暂无评论

Machine learning for inferring animal behavior from location and movement data

引用

ECOLOGICAL INFORMATICS 2019年 49卷 69-76页

作者： Wang, Guiming Mississippi State Univ Dept Wildlife Fisheries & Aquaculture Mail Stop 9690 Mississippi State MS 39762 USA

Movement ecology has rapidly advanced owing to recent developments of animal-attached devices and wide applications of sophisticated statistical and machine learning techniques in analysis of animal movement data. Global Positioning System (GPS) transmitters used for estimating animal locations and tri-axial accelerometers used for measuring the 3-dimensional accelerations of animal's motion aid researchers in collecting location and locomotion data at fine spatial and temporal scales. Machine learning and other advanced statistical methods bridge conceptual models to data, providing insights into ecological and physiological mechanisms underlying animal behavior and movements. This study reviews the general principles and applications of state space models, hidden Markov models, random forests, and support vector machines in the inference of animal behavior from movement data. Unsupervised learning algorithms, including Bayesian state space models implemented for robust correlated random walk models and hidden Markov models, help infer different behavioral modes using GPS location data. State space models can account for measurement error in GPS locations and estimate the true locations of animals;however, without including movement-state switching, state space models do not infer behavioral modes directly. On the contrary, hidden Markov models estimate the probabilities that animals switch between different behavioral modes. Nevertheless, hidden Markov models neither directly estimate animal locations nor account for measurement error explicitly. supervised learning algorithms integrate data on locations and directional accelerations with synchronized behavioral observations (i.e., labels) to classify behaviors to pre-defined behavioral categories. Unlike unsupervised learning, supervised learning requires behavioral observations to label locations and accelerometer data to train the learning algorithms. However, behavioral observations synchronized with relocatio

关键词： Behavioral classification Computational ecology Hidden Markov model Movement ecology State space model supervised learning algorithm Unsupervised learning algorithm

来源：评论

学校读者我要写书评

暂无评论

The ensemble approach in comparison with the diverse feature selection techniques for estimating NPPs parameters using the different learning algorithms of the feed-forward neural network

引用

NUCLEAR ENGINEERING AND TECHNOLOGY 2021年第12期53卷 3944-3951页

作者： Moshkbar-Bakhshayesh, Khalil Sharif Univ Technol Dept Energy Engn Azadi Ave Tehran Iran

Several reasons such as no free lunch theorem indicate that there is nota universal Feature selection (FS) technique that outperforms other ones. Moreover, some approaches such as using synthetic dataset, in presence of large number of FS techniques, are very tedious and time consuming task. In this study to tackle the issue of dependency of estimation accuracy on the selected FS technique, a methodology based on the heterogeneous ensemble is proposed. The performance of the major learning algorithms of neural network (i.e. the FFNN-BR, the FFNN-LM) in combination with the diverse FS techniques (i.e. the NCA, the F-test, the Kendall's tau, the Pearson, the Spearman, and the Relief) and different combination techniques of the heterogeneous ensemble (i.e. the Min, the Median, the Arithmetic mean, and the Geometric mean) are considered. The target parameters/transients of Bushehr nuclear power plant (BNPP) are examined as the case study. The results show that the Min combination technique gives the more accurate estimation. Therefore, if the number of FS techniques is m and the number of learning algorithms is n, by the het-erogeneous ensemble, the search space for acceptable estimation of the target parameters may be reduced from n x m to n x 1. The proposed methodology gives a simple and practical approach for more reliable and more accurate estimation of the target parameters compared to the methods such as the use of synthetic dataset or trial and error methods. (c) 2021 Korean Nuclear Society, Published by Elsevier Korea LLC. This is an open access article under the CC BY-NC-ND license (http://***/licenses/by-nc-nd/4.0/).

关键词： Combination technique Features selection technique Heterogeneous ensemble Parameter estimation supervised learning algorithm

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：