For training of process neural networks based on the orthogonal basis expansion, it is difficult to converge for BP algorithm as more parameters. Aiming at the issue, this paper proposes a solution based on quantum ge...
详细信息
ISBN:
(纸本)9780769535708
For training of process neural networks based on the orthogonal basis expansion, it is difficult to converge for BP algorithm as more parameters. Aiming at the issue, this paper proposes a solution based on quantum genetic algorithm with double chains. Firstly, the number of genes is determined by the number of weight parameters, quantum chromosomes are constructed by qubits, and the current optimal chromosome is obtained with the help of colony assessment. Secondly, taking each qubit in this optimal chromosome as the goal, individuals are updated by quantum rotation gate, and mutated by quantum non-gate to increase the diversity of population. In this method, each chromosome carrying two chains of genes, therefore it can extend ergodicity for solution space and accelerate optimization process. Taking the pattern classification of two groups of two-dimensional trigonometric functions as an example, the simulation results show that the method not only has fast convergence, but also good optimization ability.
This paper introduces a new concept of the connection weight to the multi-layer feedforward neural network. The architecture of the proposed approach is the same as that of the original multi-layer feedforward neural ...
详细信息
ISBN:
(纸本)9783642015069
This paper introduces a new concept of the connection weight to the multi-layer feedforward neural network. The architecture of the proposed approach is the same as that of the original multi-layer feedforward neural network. However, the weight of each connection is multi-valued, depending oil the value of the input data involved. The backpropagation learning algorithm was also modified to suit the proposed concept. This proposed model has been benchmarked against the original feedforward neural network and the radial basis function network. The results on six benchmark problems are very encouraging.
A Volterra equalizer based on MBER (Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear channels. Restarted BFGS could quicken convergence speed in MBER equalizer ...
详细信息
ISBN:
(纸本)9781424448203
A Volterra equalizer based on MBER (Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear channels. Restarted BFGS could quicken convergence speed in MBER equalizer trainings, and the updated matrix of BFGS is restarted conditionally and it follows that the new method becomes much more robust. By canceling line search, it is convenient to implement the new method online. In simulations, Volterra equalizers based on minimum mean square error principle degenerate rapidly in nonlinear channels, but that based on MBER provide very low bit error rate. MBER equalizers are trained online by restarted BFGS algorithm, and the results show that its convergence rate is much faster than that of stochastic gradient algorithm.
We propose a stochastic gradient descent algorithm for learning the gradient of a regression function from random samples of function values. This is a learning algorithm involving Mercer kernels. By a detailed analys...
详细信息
We propose a stochastic gradient descent algorithm for learning the gradient of a regression function from random samples of function values. This is a learning algorithm involving Mercer kernels. By a detailed analysis in reproducing kernel Hilbert spaces, we provide some error bounds to show that the gradient estimated by the algorithm converges to the true gradient, under some natural conditions on the regression function and suitable choices of the step size and regularization parameters. (C) 2007 Elsevier Inc. All rights reserved.
Considering that inputs of a process neural network(PNN) are generally time-varying functions while the inputs of many practical problems are discrete values of multiple series,in this paper,a process neural network w...
详细信息
Considering that inputs of a process neural network(PNN) are generally time-varying functions while the inputs of many practical problems are discrete values of multiple series,in this paper,a process neural network with discrete inputs is presented to provide improved forecasting results for solving the complex time series *** presented method first makes discrete input series carry out Walsh transformation,and submits the transformed series to the network for *** can solve the problem of space-time aggregation operation of *** order to examine the effectiveness of the presented method,the actual data of sunspots during 1749-2007 are *** predict the number of sunspots,the suitability of the developed model is examined in comparison with the other models to show its superiority and be an effective way of improving forecasting accuracy of networks.
A Volterra equalizer based on MBER(Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear *** BFGS could quicken convergence speed in MBER equalizer trainings, and th...
详细信息
A Volterra equalizer based on MBER(Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear *** BFGS could quicken convergence speed in MBER equalizer trainings, and the updated matrix of BFGS is restarted conditionally and it follows that the new method becomes much more *** canceling line search,it is convenient to implement the new method *** simulations,Volterra equalizers based on minimum mean square error principle degenerate rapidly in nonlinear channels,but that based on MBER provide very low bit error rate. MBER equalizers are trained online by restarted BFGS algorithm,and the results show that its convergence rate is much faster than that of stochastic gradient algorithm.
In this paper,a novel function series expansion method based on functional network model is proposed,and a functional network model for functions in several variables series expansion and learning algorithm are given,...
详细信息
In this paper,a novel function series expansion method based on functional network model is proposed,and a functional network model for functions in several variables series expansion and learning algorithm are given,the learning of parameters of the functional networks is carried out by the solving linear *** simulation results show that the proposed approach is more efficient and feasible in function series *** this algorithm,we only need the sample space of the original *** this algorithm has the value of application in industries.
Forecasting pests emergence levels plays a significant role in regional crop planting and *** accuracy,which is derived from the accuracy of the forecasting approach used,will determine the economics of the operation ...
详细信息
Forecasting pests emergence levels plays a significant role in regional crop planting and *** accuracy,which is derived from the accuracy of the forecasting approach used,will determine the economics of the operation of the pests *** methods including time series,regression analysis or ARMA model entail exogenous input together with a number of *** use of neural networks has been shown to be a cost-effective *** their training,usually with back-propagation algorithm or other gradient algorithms,is featured with some drawbacks such as very slow convergence and easy entrapment in a local *** paper presents a hybrid approach of neural network with particle swarm optimization for developing the accuracy of *** approach is applied to forecast Alternaria alternate Keissl emergence level of the WuLong Country,one of the most important tobacco planting areas in *** ARMA model and BP neural network are investigated as comparison *** experimental results show that the proposed approach can achieve better prediction performance.
A Volterra equalizer based on MBER (Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear channels. Restarted BFGS could quicken convergence speed in MBER equalizer ...
详细信息
A Volterra equalizer based on MBER (Minimum Bit Error Rate) and restarted BFGS method is proposed in this paper for equalization of nonlinear channels. Restarted BFGS could quicken convergence speed in MBER equalizer trainings, and the updated matrix of BFGS is restarted conditionally and it follows that the new method becomes much more robust. By canceling line search, it is convenient to implement the new method online. In simulations, Volterra equalizers based on minimum mean square error principle degenerate rapidly in nonlinear channels, but that based on MBER provide very low bit error rate. MBER equalizers are trained online by restarted BFGS algorithm, and the results show that its convergence rate is much faster than that of stochastic gradient algorithm.
In order to improve solving Support Vector Machine algorithm, an improved learning algorithm of the parallel SMO is proposed. According to this algorithm, the master CPU averagely distributes primitive training set to...
详细信息
In order to improve solving Support Vector Machine algorithm, an improved learning algorithm of the parallel SMO is proposed. According to this algorithm, the master CPU averagely distributes primitive training set to slave CPUs so that they can almost independently run serial SMO on their respective training set As it adopts the strategies of buffer and shrink, the speed of the parallel training algorithm is increased, which is showed in the experiments of parallel SMO based on the dataset of MNIST. The experiments indicate that the parallel SMO algorithm has good performance in solving large-scale SVM.
暂无评论