检索结果-内蒙古大学图书馆

IEEE International Conference on Industrial Technology (ICIT)

作者： A.K. Srivastava S.K. Srivastava K.K. Shukla Centre For Research In Microelectronics Department of Electronics EngineeringInstitute of Technology Banaras Hindu University Varanasi India Visual Computing and Artificial Intelligence Laboratories Department of Computer EngineeringInstitute of Technology Banaras Hindu University Varanasi India

This paper reports the effect of some advanced genetic operators like two-parents multipoint restricted crossover (Double-MRX), three-parents multipoint restricted crossover (Triple-MRX), elitist selection and scheduled mutation on the adaptability of feedforward neural networks trained over complex and computationally expensive electronic nose data. The authors show that the performance of Triple-MRX is better that of Double-MRX. Upon applying elitist selection with Double-MRX and scheduled mutation with Triple-MRX, the performance of the genetic training of the neural network improves up to some extent, but Triple-MRX is still better than Double-MRX as far as quality of solution and speed of convergence are concerned. It is also shown that the performance levels of these hybrid techniques far exceeds those of the commonly used backpropagation model. The search for a good neuro-genetic hybrid computational paradigm based on advanced genetic operators is a frontier research area in the evolution of a sixth generation computing system.

关键词： Neural networks Genetic mutations Genetic algorithms Electronic mail Artificial intelligence Convergence Biological cells Sensor arrays Genetic engineering Network topology

来源：评论

学校读者我要写书评

暂无评论

On the performance evaluation of hybrid-trained neural classifier for the detection of hazardous vapours using responses from SAW sensors array

On the performance evaluation of hybrid-trained neural class...

引用

IEEE International Conference on Industrial Technology (ICIT)

作者： A.K. Srivastava S.K. Srivastava K.K. Shukla Centre For Research In Microelectronics Department of Electronics EngineeringInstitute of Technology Banaras Hindu University Varanasi India Visual Computing and Artificial Inlelligence Laboratories Department of Computer EngineeringInstitute of Technology Banaras Hindu University Varanasi India

A neural classifier has been designed by a new two-phase hybrid training algorithm introduced by us for classification of hazardous vapours. The neural network is trained using genetic algorithm in initial phase. This is followed by a second phase of backpropagation training that uses weight matrix determined by first phase for initialization. For establishing the superior performance of our classifier, published data from polymer-coated surface-acoustic wave (SAW) sensors array exposed to varying concentration of each of nine vapours belonging to two different classes have been used. Vapours of class I are toxic vapours of interest in ambient air that contains common interferents (class II vapours) at much higher concentration. Performance of the classifier is evaluated by reducing dimensionality of resulting data matrix from 4 to 1 by taking a different set of sensors. We show that as the dimension is reduced, the gas identification problem becomes harder for backpropagation. Whereas the same set of problems when solved using a genetic algorithm with heuristic switch over to backpropagation as a training paradigm, significantly better results are obtained in predicting class and type of test vapours.

关键词： Sensor arrays Surface acoustic waves Genetic algorithms Switches Algorithm design and analysis Backpropagation algorithms Neural networks Polymers Surface waves Testing

来源：评论

学校读者我要写书评

暂无评论

On the design issue of intelligent electronic nose system

On the design issue of intelligent electronic nose system

引用

IEEE International Conference on Industrial Technology (ICIT)

作者： A.K. Srivastava S.K. Srivastava K.K. Shukla Centre For Research In Microelectronics Departnient of Electronics Engineering Institute of Technology Banaras Hindu University Varanasi India Visual Computing and Artificial Intelligence Laboratories Department of Computer Engineering Institute of Technology Banaras Hindu University Varanasi India

Intelligent electronic nose (ENOSE) system technology is gaining importance in several industrial applications. These include process control and quality control in industries such as foodstuffs, beverages, tobacco, perfumery and pharmaceutical. ENOSE is also crucial component in industrial safety (smoke and hazardous gas detection) as well as environmental pollution control. This paper deals with design of an intelligent ENOSE system for the identification of gas/odours using a sensor array and a neural network pattern classifier. Previous researchers have shown that the power of discrimination increases rapidly with the number of sensors in the array whose information potential is very large and the pattern recognition (PARC) method is a clever way to extract this information. The authors show in this paper with the powerful PARC technique, the need of larger array can be compensated. With this view, they design a neural classifier using two different learning approaches and train the network over the responses of surface acoustic wave (SAW) sensors exposed to hazardous vapours like diethyl sulphide (DES) and iso-octane (ISO). Dimensionality of the data set is varied from 1 to 8 by taking different number of sensors. It is found that for a backpropagation trained neural classifier, the optimum number of sensors required for satisfactory classification under noisy conditions is 4 to 5. This is a very limited range beyond which backpropagation has great difficulty in training the neural classifier even with repeated restarts and different weight initializations. To alleviate this problem, hybridization of soft computing tools like neural networks and genetic algorithms promises to provide the design of better intelligent system. The authors propose the use of a genetic algorithm based on a special MRX operator introduced by them and demonstrate very encouraging results with genetically trained neural network model even with larger as well as smaller numbers of senso

关键词： Intelligent systems Electronic noses Sensor arrays Intelligent sensors Electrical equipment industry Neural networks Acoustic sensors Industrial control Intelligent networks Surface acoustic waves

来源：评论

学校读者我要写书评

暂无评论

A new model of global parallel genetic algorithm

A new model of global parallel genetic algorithm

引用

International Conference on Information Technology Interfaces

作者： M. Golub D. Jakobovic Fac. of Electr. Eng. & Comput. Zagreb Univ. Croatia Department of Electronics Microelectronics Computer and Intelligent SystemsFaculty of Electrical Engineering and Computing University of Zagreb Zagreb Croatia

ISBN: (纸本)9539676916

The authors describe a multithreaded parallel genetic algorithm (PGA) implementation. Considering the basic models of parallel genetic algorithms, we identify a variant of global PGA (GPGA) as the most appropriate one for use on a multiprocessor system with few processors. The difference between the synchronous and asynchronous model is analyzed and their characteristics are evaluated. Unlike some authors, we choose not to allow a single individual to be engaged in a tournament competition in more than one instance (no duplicates). The probability of selection for elimination of an individual is than determined based on the fitness of the chromosome and compared with the same probability of the duplicate-allowing algorithm. Finally, the main advantages and disadvantages of the GPGA as well as performance comparison with sequential GA are stated.

关键词： Genetic algorithms electronics packaging Concurrent computing Master-slave Multithreading Space exploration microelectronics Intelligent systems Multiprocessing systems Yarn

来源：评论

学校读者我要写书评

暂无评论

Frequency domain analysis of B-spline interpolation

Frequency domain analysis of B-spline interpolation

引用

IEEE International Symposium on Industrial electronics (ISIE)

作者： Z. Mihajlovic A. Goluban M. Zagar Department of Electronics Microelectronics Computer and Intelligent Systems Faculty of Electrical Engineering and Computing University of Zagreb Zagreb Croatia Department of Control and Computer Engineering in Automation Faculty of Electrical Engineering and Computing University of Zagreb Zagreb Croatia

This paper describes B-spline interpolation and compares it with other reconstruction methods, especially in three-dimensional space. We first consider the B-spline bases in the terms of convolution in signal processing. The presented analysis requires careful usage of continuous and discrete representation of B-spline. Emphasis is given to the important difference between B-spline interpolation and approximation. The difference is shown through frequency domain analysis, so we derive frequency responses of the B-spline interpolation and approximation. We conclude by demonstrating the use of several reconstruction filters and appropriate gradient estimators in volume rendering. Exact reconstruction in volume visualization is very important in many industrial applications, such as material cavity control.

关键词： Frequency domain analysis Spline Interpolation Image reconstruction Convolution Filters Polynomials Data visualization Frequency response microelectronics

来源：评论

学校读者我要写书评

暂无评论

The B-spline interpolation in visualization

引用

Journal of Computing and Information Technology 1999年第3期7卷 245-253页

作者： Mihajlović, Željka Goluban, Alan Department of Electronics Microelectronics Computer and intelligent Systems Faculty of Electrical Engineering arid Computing University of Zagreb Unska 3 Zagreb10 000 Croatia Department of Control and Computer Engineering in Automation Faculty of Electrical Engineering and Computing University of Zagreb Unska 3 Zagreb10 000 Croatia

The volume data is generally in the form of the large array of numbers. In order to render the object hidden in the volumetric data, we need to reconstruct or interpolate data values between the samples. The novelty presented in this paper is B-spline interpolation in the volumetric space. We show that this approach is better than currently used methods. We also present a hybrid approach, analyze this approach in frequency domain and compare it to B-spline interpolation. To enhance the quality during the volume visualization process it is important to enhance the quality of the reconstruction. It is of crucial importance to explore different undesired effects. If better reconstruction is performed the more accurate result of volume visualization process is achieved.

关键词： Volume rendering

来源：评论

学校读者我要写书评

暂无评论

Joint optimization of quantization and on-line channel estimation for low bit-rate video transmission

Joint optimization of quantization and on-line channel estim...

引用

IEEE International Conference on Image Processing

作者： R. Chandramouli S. Kumar N. Ranganathan Center for Microelectronics Research Department of Computer Science and Engineering University of South Florida Tampa USA Department of ECE University of Texas El Paso El Paso TX USA

Optimal quantization and channel estimation are one among the main issues in low bit-rate video transmission over time varying noisy channels. Previous approaches to these issues were mainly based on quantizers optimized for parametric channel models and pilot symbol aided techniques for channel identification. However, this optimality may not hold when the randomly varying channel behavior deviates from these models. Also, the cost involved in terms of delay could be large for pilot symbol based channel estimation. We propose a new empirically optimized channel matched quantizer and a stochastic learning algorithm that estimates and tracks the channel with minimal additional delay and overhead. Performance analysis of the algorithm shows that the new adaptive quantizer results in a better quality video. The learning algorithm converges very fast.

关键词： Quantization Channel estimation Bit error rate Video compression Delay estimation Stochastic processes PSNR Learning automata Discrete cosine transforms microelectronics

来源：评论

学校读者我要写书评

暂无评论

A residue number arithmetic based circuit for pipelined computation of autocorrelation coefficients of speech signal

A residue number arithmetic based circuit for pipelined comp...

引用

International Conference on VLSI Design

作者： A. Drolshagen C.C. Sekhar W. Anheier Institute for Theoretical Electronics and Microelectronics (ITEM) FBI University of Bremen Bremen Germany Department of Computer Science and Engineering Indian Institute of Technology Chennai India

This paper presents a new design for module multipliers suitable for moduli that are not necessarily prime. The design avoids the need for building special purpose look-up tables. An RNS arithmetic based autocorrelator circuit that uses a large number of these multipliers is designed. The paper also describes the automated design strategy using a hardware compiler for RNS, and presents the results of the synthesis.

关键词： Arithmetic Circuits Autocorrelation Speech analysis Signal processing Signal design Speech synthesis Signal analysis Speech processing Hardware

来源：评论

学校读者我要写书评

暂无评论

1-wavelength loop type dielectric chip antennas

1-wavelength loop type dielectric chip antennas

引用

Antennas and Propagation Society International Symposium

作者： H. Tanidokoro N. Konishi E. Hirose Y. Shinohara H. Arai N. Goto Microelectronics Research Laboratory Central Research Institute Mitsubishi Material Corporation Chichibu Saitama Japan Diviaion of Electrical and Computer Engineering Faculty fo Engineering Yokohama National University Yokohama Kanagawa Japan Department of Electronics and Systems. Faculty of Engineering Takushoku University Hachioji Tokyo Japan

We propose a 1-wavelength loop type dielectric chip antenna that meets the demand for antennas with unidirectional radiation patterns for portable communication devices. The structure of the proposed antenna is shown. The radiating element printed loop is on the top of the cube. A feeding strip line is on one of the vertical face of the cube and a shorted plate parallel to the feeding strip line is on the same face. The conductive plane is on the bottom of the cube. By printing all of conductive elements on a ceramic cube this novel antenna is easy to fabricate.

关键词： Ceramics Finite difference methods Time domain analysis Dielectric loss measurement Directive antennas Antenna radiation patterns Conductors Permittivity Thick films Antenna measurements

来源：评论

学校读者我要写书评

暂无评论

Automatic Generation of Modular Time-Space Mappings and Data Alignments

引用

Journal of VLSI Signal Processing Systems for Signal, Image, and Video Technology 1998年第2期19卷 195-208页

作者： Lee, Hyuk-Jae Fortes, José A.B. Department of Computer Science Louisiana Tech. University Ruston LA 71272 United States School of ECE Purdue University W. Lafayette IN 47907 United States Seoul National University Seoul South Korea University of Arizona Tucson AZ United States Purdue University West Lafayette IN United States Department of Computer Science Louisiana Tech. University Ruston LA United States Microcontroller Devmt. Syst. Grp. S. Ki-heung South Korea Universidade de Angola Angola Colorado State University Fort Collins CO United States University of Southern California Los Angeles CA United States School of Electrical Engineering Purdue University West Lafayette IN United States Computer Architecture Department Universitat Politecnica de Catalunya Barcelona Spain Office of Naval Research AT and T Foundation Gen. Elec. and the Natl. Sci. Found. United States Inst. Elec. Electronics Engineers United States

Time-space transformations and data alignments that can lead to efficient execution of parallel programs have been extensively studied. Recently, modular time-space transformations have been proposed to generate a class of algorithm mappings that cannot be described by linear time-space transformations. This paper proposes a new class of data alignments, called expanded modular data alignments (EMDAs), for programs that result from modular time-space transformations. An EMDA subsumes multiple modular data alignments, which are described by affine functions modulo a constant vector. Conditions of a modular time-space mapping and an EMDA for perfect alignment are described. However, these conditions together with other conditions for validity and optimality of a modular mapping introduce nonlinear constraints in the problem of generating modular mappings. A method of O (n ²) complexity is provided to choose some entries of a transformation matrix so that nonlinear constraints are transformed into linear ones, where n is the dimension of the computation domain (e.g., the number of nested loops). Although the solution space of the problem is reduced by assigning fixed values to some entries, the proposed heuristic attempts to minimize the number of the fixed entries and consequently to exclude as few solutions as possible.

关键词：

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：