In this paper, the bipartite graphical structure of LDPC codes is studied, some methodology for designing shorten LDPC Codes are presented. The performance evaluation of shorten LDPC codes presented provides easy and ...
详细信息
ISBN:
(纸本)9783038350194
In this paper, the bipartite graphical structure of LDPC codes is studied, some methodology for designing shorten LDPC Codes are presented. The performance evaluation of shorten LDPC codes presented provides easy and effective strategy in selecting appropriate shorten scheme, avoiding the complex encoding and iterative decoding steps. The shorten LDPC codes whose minimum code weight greater and length 4-cycle free have better performance.
This paper proposes a new approach for training FNN by hybrid DE and *** combines the advantages of the global search performed by DE over the FNN parameter space and the local search of *** a function approximation a...
详细信息
This paper proposes a new approach for training FNN by hybrid DE and *** combines the advantages of the global search performed by DE over the FNN parameter space and the local search of *** a function approximation as an illustration,we compare the HDEbp and bp for effectiveness and efficiency for training *** shows that the use of new method can provide better results than bp.
It is one of the most important tasks to improve the training time in the back-propagation(bp) neural *** this paper two new methods based on error back propagation by adopting dynamic adjusting weights for reduction ...
详细信息
It is one of the most important tasks to improve the training time in the back-propagation(bp) neural *** this paper two new methods based on error back propagation by adopting dynamic adjusting weights for reduction the training time are *** approaches are based on an adequate modification of the traditional and classical *** interesting results of computer experiments with the modified bp algorithm are *** results prove that these new methods are effective to solve some problems and faster than the traditional methods in training multi-layer feed-forward neural networks.
In order to improve the prediction ability of stock price, a prediction method based on Wavelet Neural Network(WNN) is proposed. First bp algorithm is used to optimize the parameters of WNN with Mexico Hat wavelet bas...
详细信息
In order to improve the prediction ability of stock price, a prediction method based on Wavelet Neural Network(WNN) is proposed. First bp algorithm is used to optimize the parameters of WNN with Mexico Hat wavelet basis for the establishment of the stock price prediction model, and then the built model is applied to predict the stock price movement on the basis of 15 features. The simulations on daily closing price index of SSE Composite Index indicate that, the proposed method has the advantages of simple structure, strong implementation and good prediction accuracy, and gets better stock price prediction in contrast with single neural network and genetic neural network. This verifies the feasibility and effectiveness of the method in the application of stock price prediction.
In this paper, we propose a modification to Belief Propagation (bp) decoding algorithm for LDPC codes. The modification is to attenuate the check to bit extrinsic logarithm likelihood ratio by a factor a, when sudden ...
详细信息
ISBN:
(纸本)9781424435746
In this paper, we propose a modification to Belief Propagation (bp) decoding algorithm for LDPC codes. The modification is to attenuate the check to bit extrinsic logarithm likelihood ratio by a factor a, when sudden sign change happens. This modification can be applied to both the standard bp algorithm and the joint row and column (JRC) bp algorithm. Simulation results show that the BER and WER performance of both traditional bp and JRC bp algorithms is improved by this method. The expense of the proposed modification is a slight increase in the average number of decoding iterations.
Based on fuzzy mathematics and the neural network method, with the combination of fault datas of B737NG aircraft flight control system and experts experiences, this paper built a fuzzy neural network (FNN) fault diagn...
详细信息
Based on fuzzy mathematics and the neural network method, with the combination of fault datas of B737NG aircraft flight control system and experts experiences, this paper built a fuzzy neural network (FNN) fault diagnosis model which can intuitively express the relevance of fault symptoms and fault reasons. By Simulation and analysis, the fault diagnosis accuracy of FNN is much higher than only bp neural network, The research makes fault positioning more accurate and valid, improves the efficiency of maintenance, as well as ensures flight safety and provides a new point of view for establishing the aviation maintenance auxiliary decision system. (C) 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of [CEIS 2011]
By using alternating current plasma arc welding,the influences were studied of such parameters as welding curent,arc voltage,welding speed,wire feed rate,and magnitude of ion gas flow on front melting width,wdle rei...
详细信息
By using alternating current plasma arc welding,the influences were studied of such parameters as welding curent,arc voltage,welding speed,wire feed rate,and magnitude of ion gas flow on front melting width,wdle reinforcement,and back melting width of LF6 aluminum *** of the formation of welding seam in alternating current plasma arc welding of aluminum was set up with the method of artificial neural neural network - bp algorithm. Qyakuty of formation was consequently predicted and *** experimental result shows that,compared with other modeling methods,artificial network model can be used to more accurately predict formation of weld,and to guide the production practice.
This paper is about the application of ANN (artificial neural networks) theory in evaluation of mine design schemes and a quantified evaluation method based on a three\|layer neural network is given. It studies the st...
详细信息
This paper is about the application of ANN (artificial neural networks) theory in evaluation of mine design schemes and a quantified evaluation method based on a three\|layer neural network is given. It studies the structure of the three\|layer neural network, its learning process, its operating algorithm to realize the evaluation of mine design schemes in a computer and a practical example is also involved in it.
The standard back-propagation (bp) algorithm for multilayer feedforward neural networks is basically a gradient-descent method, it has the problems of local minima and slow convergence. In this paper, a simple method ...
详细信息
This paper provides a character recognition system based on the Back Propagation artificial neural network that is *** with samples of the characters,the characters are recognized,using with VC++.Recognition rate reac...
详细信息
This paper provides a character recognition system based on the Back Propagation artificial neural network that is *** with samples of the characters,the characters are recognized,using with VC++.Recognition rate reaches 96%,which meets the design requirement.
暂无评论