检索结果-内蒙古大学图书馆

International Joint Conference on Neural Networks (IJCNN)

作者： Wakamatsu, Toshiki Takase, Haruhiko Kawanaka, Hiroharu Tsuruoka, Shinji Mie Univ Grad Sch Engn Tsu Mie 514 Japan

ISBN: (纸本)9781424496365

In this paper, we aims to improve stability of learning processes by the SpikeProp algorithm. We proposed the method that reduce the increase of the error in learning processes. It repeats two steps: (1) original SpikeProp algorithm, and (2) use a linear search in the steepest descent direction only if the first step is failed. Some experimental results shows the improvement of learning processes.

关键词： Biological neural networks Iris Iris recognition Mathematical model Neurons SpikeProp algorithm SpikeProp network Surges training feed-forward networks feedforward neural nets learning (artificial intelligence) learning process spiking neural networks stability stability improvement supervised learning algorithm training algorithm

来源：评论

学校读者我要写书评

暂无评论

A training method for SpikeProp without redundant spikes -Removing unnecessary sub-connections during training- 12

A training method for SpikeProp without redundant spikes -Re...

引用

12th International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery (ICNC-FSKD)

作者： Nakayama, Takutoshi Matsumoto, Takashi Takase, Haruhiko Kawanaka, Hiroharu Tsuruoka, Shinji Mie Univ Grad Sch Engn Tsu Mie 5148507 Japan Mie Univ Tsu Mie 5148507 Japan

ISBN: (纸本)9781509040933

SpikeProp, which is proposed by Bohte and extended by Booij, is a type of multi-layer networks of spiking neurons. Our research group has proposed a training algorithm for SpikeProp without redundant output spikes. However, its performance is sensitive to the initial network structure, such as the number of hidden units, the number of sub-connections, and their delays. In this article, we discuss the problem, especially the number of sub-connections. And we propose a method to remove unnecessary sub-connections during training. By simple experiments, we show that the method brings less dependency of training performance on the number of initial sub-connections.

关键词： Spiking Neural Network SpikeProp training algorithm Time Series Signal Processing

来源：评论

学校读者我要写书评

暂无评论

Bandwidth Enhancement by Direct Coupled Antenna for WLAN/GPS/WiMax Applications and Feed Point Coordinate Analysis through ANN

引用

WIRELESS PERSONAL COMMUNICATIONS 2016年第1期91卷 9-32页

作者： Ayub, Shahanaz Srivastava, Rajat Bundelkhand Inst Engn & Technol Jhansi Uttar Pradesh India

In the present work first of all the bandwidth of rectangular Microstrip antenna is enhanced by direct coupling at the operating frequency 2.4 GHz using IE3D software and then it is analyzed through IE3D simulation software for maximum bandwidth for different antenna dimensions at different operating frequencies (ranging from 1.2 to 2.5 GHz). The rectangular line feed is used to energize the proposed antenna (at 2.4 GHz operating frequency), the feed point coordinate is varied as the antenna dimensions varied so it is required to analyze the antenna for the feed point coordinates where it provides the maximum bandwidth. The simulation of antenna for its maximum bandwidth through most of the antenna simulation software's is a time taking process. In order to reduce this time, the Artificial Neural Network (ANN) approach is adopted here. For this purpose a Feed Forward Neural Network with Levenberg-Marquardt training algorithm is used. Determining the feed point coordinates through proposed ANN takes only 1-2 s where as antenna simulation softwares (like IE3D) take time in hours. The proposed antenna has dual frequency band. In first frequency band (1.392-1.44 GHz) the fractional bandwidth is 3.38 % and in second frequency band (1.733-3.58 GHz) the fractional bandwidth is 69.5 % which is suitable for broad band applications.

关键词： Direct coupling Bandwidth enhancement Artificial neural network training algorithm Feed forward networks Gain Feed point coordinates IE3D

来源：评论

学校读者我要写书评

暂无评论

A Novel Quasi-Newton-Based training Using Nesterov's Accelerated Gradient for Neural Networks 25th

A Novel Quasi-Newton-Based Training Using Nesterov's Acceler...

引用

25th International Conference on Artificial Neural Networks (ICANN)

作者： Ninomiya, Hiroshi Shonan Inst Technol Dept Informat Sci Fujisawa Kanagawa Japan

ISBN: (纸本)9783319447810;9783319447803

Neural networks have been recognized as a useful tool for the function approximation problems with high-nonlinearity [1]. training is the most important step in developing a neural network model. Gradient based algorithms are popularly used for the training. The popular methods are first-order methods so called Steepest Gradient (SG), Classical Momentum (CM), Nesterov’s Accelerated Gradient (NAG) methods and a more dedicated method such as RPROP. On the other hand, quasi-Newton (QN) method which is one of the most efficient optimization techniques with super-linear convergence is also widely utilized for more complicated problems. However, when applied to highly nonlinear function modeling, the above methods even QN still converges too slowly and optimization error cannot be effectively reduced within finite time. This paper describes a novel acceleration technique of QN using Nesterov’s accelerated gradient [2]. The proposed algorithm is referred to as Nesterov’s accelerated quasi-Newton (NAQ) method. The update equation of the approximated matrix of Hessian is mathematically derived from the quadratic approximation of error function. The Hessian is approximated using both of the normal gradient and the Nesterov’s accelerated one in NAQ. NAQ is a technique for accelerating QN that accumulates an update vector in directions of persistent reduction in the objective across iterations. As a result, the iteration during the NAQ training can be shortened without loss of the strong ability to search a global minimum in QN. The proposed algorithm is demonstrated through the computer simulations for the benchmark problems with high-nonlinearity compared with SG, CM, NAG, RPROP and QN. From the simulation results it is shown that NAQ is faster than the conventional methods without compromising quality of training solutions. © Springer International Publishing Switzerland 2016.

关键词： Neural networks training algorithm Nesterov's accelerated gradient method Quasi-Newton method

来源：评论

学校读者我要写书评

暂无评论

Operational parameter impact and back propagation artificial neural network modeling for phosphate adsorption onto acid-activated neutralized red mud

引用

JOURNAL OF MOLECULAR LIQUIDS 2016年第0期216卷 35-41页

作者： Ye, Jie Cong, Xiangna Zhang, Panyue Zeng, Guangming Hoffmann, Erhard Wu, Yan Zhang, Haibo Fang, Wei Hunan Univ Coll Environm Sci & Engn Changsha 410082 Hunan Peoples R China Hunan Univ Minist Educ Key Lab Environm Biol & Pollut Control Changsha 410082 Hunan Peoples R China Karlsruhe Inst Technol Dept Aquat Environm Engn D-76131 Karlsruhe Germany Karlsruhe Inst Technol Inst Appl Mat IAM WK D-76131 Karlsruhe Germany

In this research the combination of neutralization activation and acid activation processes was employed to improve the physicochemical characters of red mud. In order to better understand the phosphate adsorption behaviors and further improve the phosphate adsorption performance of acid-activated neutralized red mud (AaN-RM), for the first time the impact of operational parameters on phosphate adsorption onto AaN-RM was systematically investigated, and back propagation artificial neural network (ANN) modeling was conducted. The results demonstrated that phosphate adsorption capacity of AaN-RM decreased with the enhancement of adsorbent dosage and the concentration of the competing ion (carbonate), while it increased with the increase of initial phosphate concentration and contact time. The optimal adsorption temperature and initial solution pH for phosphate adsorption onto AaN-RM were 50 degrees C and 4.0, respectively. Moreover, a 6-10-1 feed forward ANN structure with trainlm algorithm was successfully constructed for predicting the phosphate removal by AaN-RM. The RMSE and R-2 values for two subsets (training and validation subset, and testing subset) were 3.06 and 2.61, and 0.9932 and 0.9969, respectively. Furthermore, the importance analysis showed that contact time and initial phosphate concentration were the most influential parameters on phosphate removal by AaN-RM, the importance of which reached 24.64% and 22.16%, respectively. (C) 2016 Elsevier B.V. All rights reserved.

关键词： Phosphate adsorption Red mud Artificial neural network training algorithm Importance analysis

来源：评论

学校读者我要写书评

暂无评论

Wavelet Neural Network with Random Wavelet Function Parameters

引用

INTERNATIONAL JOURNAL OF ENGINEERING 2017年第10期30卷 1510-1516页

作者： Bazoobandi, H. Esfarayen Univ Technol Dept Comp Engn Esfarayen North Khorasan Iran

The training algorithm of Wavelet Neural Networks (WNN) is a bottleneck which impacts on the accuracy of the final WNN model. Several methods have been proposed for training the WNNs. From the perspective of our research, most of these algorithms are iterative and need to adjust all the parameters of WNN. This paper proposes a one-step learning method which changes the weights between hidden layer and output layer of the network;meanwhile, the wavelet function parameters are randomly assigned and kept fixed during the training process. Besides the simplicity and speed of the proposed one-step algorithm, the experimental results verify the performance of the proposed method in terms of final model accuracy and computational time.

关键词： Wavelet Neural Network training algorithm Moore-Penrose Inverse Random Parameter Values

来源：评论

学校读者我要写书评

暂无评论

Application of Artificial Neural Networks for Modeling Drug Release from a Bicomponent Hydrogel System 20

Application of Artificial Neural Networks for Modeling Drug ...

引用

20th International Conference on System Theory, Control and Computing (ICSTCC)

作者： Dumitriu, Tiberius Dumitriu, Raluca Petronela Manta, Vasile Gheorghe Asachi Tech Univ Iasi Fac Automat Control & Comp Engn Iasi Romania Petru Poni Inst Macromol Chem Phys Chem Polymers Dept Iasi Romania

ISBN: (纸本)9781509027200

Artificial Neural Networks (ANNs) have been used as modeling tools for prediction of drug release patterns from bicomponent hydrogel systems based on poly(N-isopropylacrylamide) and sodium alginate. The process modeling was performed using an artificial neural network trained with an evolutionary algorithm, the last one having the role of developing the neural model in an optimal form. The ANN was trained with this algorithm using the available experimental data as the training set. The divergence of the root mean squared error (RMSE) between the output and target values of test set was used as stop criterion. The simulation results showed that drug release profiles from the chosen hydrogels can be modeled accurately using ANNs, the model predictions being closely correlated with the experimental data.

关键词： artificial neural networks predictive modeling hydrogels drug delivery training algorithm

来源：评论

学校读者我要写书评

暂无评论

A training method for SpikeProp without redundant spikes - Removing unnecessary sub-connections during training

A training method for SpikeProp without redundant spikes - R...

引用

International Conference on Natural Computation, Fuzzy Systems and Knowledge Discovery

作者： Takutoshi Nakayama Takashi Matsumoto Haruhiko Takase Hiroharu Kawanaka Shinji Tsuruoka Graduate School of Engineering Mie university Tsu Mie 514-8507 Mie university Tsu Mie 514-8507

ISBN: (纸本)9781509040940

关键词： Spiking Neural Network SpikeProp training algorithm Time Series Signal Processing

来源：评论

学校读者我要写书评

暂无评论

Memristor Crossbar Based Unsupervised training

Memristor Crossbar Based Unsupervised Training

引用

IEEE National Aerospace and Electronics Conference (NAECON)

作者： Hasan, Raqibul Taha, Tarek M. Univ Dayton Dept Elect & Comp Engn Dayton OH 45469 USA

ISBN: (纸本)9781467375658

Several big data applications are particularly focused on classification and clustering tasks. Robustness of such system depends on how well it can extract important features from the raw data. For big data processing we are interested for a generic feature extraction mechanism for different applications. autoencoder is a popular unsupervised training algorithm for dimensionality reduction and feature extraction. In this work we have examined memiistor crossbar based implementation of autoencoder which will consume very low power. We have designed on-chip training circuitry for the unsupervised training scheme.

关键词： Low power architecture memristor crossbars autoencoder training algorithm feature extraction

来源：评论

学校读者我要写书评

暂无评论

Manhattan Rule training for Memristive Crossbar Circuit Pattern Classifiers 9

Manhattan Rule Training for Memristive Crossbar Circuit Patt...

引用

IEEE 9th International Symposium on Intelligent Signal Processing (WISP)

作者： Zamanidoost, Elham Bayat, Farnood M. Strukov, Dmitri Kataeva, Irina Univ Calif Santa Barbara Dept Elect & Comp Engn Santa Barbara CA 93106 USA Denso Corp Adv Res Div Komenoki Nisshin Japan

ISBN: (纸本)9781479972524

We investigated batch and stochastic Manhattan Rule algorithms for training multilayer perceptron classifiers implemented with memristive crossbar circuits. In Manhattan Rule training, the weights are updated only using sign information of classical backpropagation algorithm. The main advantage of Manhattan Rule is its simplicity, which leads to more compact hardware implementation and faster training time. Additionally, in case of stochastic training, Manhattan Rule allows performing all weight updates in parallel, which further speeds up the training procedure. The tradeoff for simplicity is slightly worse classification performance. For example, simulation results showed that classification fidelity on Proben1 benchmark for memristor-based implementation trained with batch Manhattan Rule were comparable to that of classical backpropagation algorithm, and about 2.8 percent worse than the best reported results.

关键词： Crossbar memory Memristor Artificial neural network training algorithm Pattern classification

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：