用于电磁兼容预测的函数链神经网络

为了解决传统BP神经网络的电磁兼容性预测方法存在易于局部收敛的问题,提出了一种基于模糊测度的函数链神经网络。该网络通过函数链将初始权重扩展到更高维度上,在实现传统BP网络多层感知的功能同时,计算过程仅为单层运算,因以收敛速度比多层的BP神经网络更快,解决了网络在解决非线性问题时,收敛于局部最小的问题。实验结果表明,提出的基于模糊测度的函数链神经网络在预测电磁兼容性方面更加精确。     

Nonlinear vector prediction using feed-forward neural networks

The performance of a classical linear vector predictor is limited by its ability to exploit only the linear correlation between the blocks. However, a nonlinear predictor exploits the higher order correlations among the neighboring blocks, and can predict edge blocks with increased accuracy. We have investigated several neural network architectures that can be used to implement a nonlinear vector predictor, including the multilayer perceptron (MLP), the functional link (FL) network, and the radial basis function (RBF) network. Our experimental results show that a neural network predictor can predict the blocks containing edges with a higher accuracy than a linear predictor.     

正则化在逻辑回归与神经网络中的应用研究

在机器学习算法的应用中,当使用小规模、多特征数的训练样本时,模型容易出现过拟合现象。正则化方法可以在一定程度上抑制模型过拟合,提高模型的泛化能力。以手写数字识别为例,分别研究了正则化方法在逻辑回归和BP神经网络中的应用,并比较了两种算法的实际结果。从MNIST手写体数据库中随机选取了5000个样本,经过PCA(Principal Component Analysis)降维后,从中选取不同规模样本并分别将其随机划分为60%的训练集,20%的交叉验证集和20%的测试集。分别构建两种算法对样本进行训练和测试,通过学习曲线选取合适的正则化参数,并比较了在合适正则化参数与未加入正则化参数下,模型与对测试集的预测精度。实验结果表明BP神经网络对手写数字的识别效果优于逻辑回归;同时当使用样本集较小时,正则化方法可以有效地抑制模型过拟合的发生,提高模型预测精度;随着样本集规模的增大,抑制效果减弱。     

深度神经网络在风电功率预测中的应用研究

风电功率预测对于电网建设具有重要意义。文中提出一种基于深度神经网络的风电功率预测方法,该方法充分考虑了影响风电功率的若干因素,如风速、风向、空气密度和季节,通过深度神经网络训练、学习给出最佳的功率预测值。通过深度神经网络学习的特征能够更为有效地反映实际情况,因此提高了风电功率预测的稳健性。基于内蒙古某风电厂的实测数据进行了验证实验,结果表明了提出方法的有效性。     

Evolutionary fuzzy neural networks for hybrid financial prediction

In this paper, an evolutionary fuzzy neural network using fuzzy logic, neural networks (NNs), and genetic algorithms (GAs) is proposed for financial prediction with hybrid input data sets from different financial domains. A new hybrid iterative evolutionary learning algorithm initializes all parameters and weights in the five-layer fuzzy NN, then uses GA to optimize these parameters, and finally applies the gradient descent learning algorithm to continue the optimization of the parameters. Importantly, GA and the gradient descent learning algorithm are used alternatively in an iterative manner to adjust the parameters until the error is less than the required value. Unlike traditional methods, we not only consider the data of the prediction factor, but also consider the hybrid factors related to the prediction factor. Bank prime loan rate, federal funds rate and discount rate are used as hybrid factors to predict future financial values. The simulation results indicate that hybrid iterative evolutionary learning combining both GA and the gradient descent learning algorithm is more powerful than the previous separate sequential training algorithm described in.     

Aperture antenna shape prediction by feedforward neural networks

The emergence of adaptive “smart” materials has led to the design of active aperture antennas. Inherent in these antennas is the ability to change their shape in real time to meet various performance characteristics. When examining the usefulness of these antennas, one of the primary concerns is the antenna shape needed for a particular radiation pattern. Aperture antenna shape prediction is also a concern in the industrial production of semi-paraboloidal antennas. The work in this study employs an artificial neural network to model the aperture antenna shape in real time. To test the accuracy of the network, the “threefold holdout technique” was employed. In this technique, sets of examples are “held out” of the training process and used to obtain the “true error” of the network. The network accurately predicted the aperture shape exactly, to within three significant digits, 96% of the time     

δ-NARMA neural networks: a new approach to signal prediction

This paper presents a new connectionist architecture for stochastic univariate signal prediction. After a review of related statistical and connectionist models pointing out their advantages and limitations, we introduce the ϵ-NARMA model as the simplest nonlinear extension of ARMA models. These models then provide the units of a MLP-like neural network: the δ-NARMA neural network. The associated learning algorithm is based on an extension of classical backpropagation and on the concept of virtual error. Such networks can be seen as an extension of ARIMA and ARARMA models and face the problem of nonstationary signal prediction. A theoretical study brings understanding of experimental phenomena observed during the δ-NARMA learning process. The experiments carried out on three railroad-related real-life signals suggest that δ-NARMA networks outperform other studied univariate models     

Environment-adaptation mobile radio propagation prediction usingradial basis function neural networks

This paper investigates the application of a radial basis function (RBF) neural network to the prediction of field strength based on topographical and morphographical data. The RBF neural network is a two-layer localized receptive field network whose output nodes from a combination of radial activation functions computed by the hidden layer nodes. Appropriate centers and connection weights in the RBF network lead to a network that is capable of forming the best approximation to any continuous nonlinear mapping up to an arbitrary resolution. Such an approximation introduces best nonlinear approximation capability into the prediction model in order to accurately predict propagation loss over an arbitrary environment based on adaptive learning from measurement data. The adaptive learning employs hybrid competitive and recursive least squares algorithms. The unsupervised competitive algorithm adjusts the centers while the recursive least squares (RLS) algorithm estimates the connection weights. Because these two learning rules are both linear, rapid convergence is guaranteed. This hybrid algorithm significantly enhances the real-time or adaptive capability of the RBF-based prediction model. The applications to Okumura's (1968) data are included to demonstrate the effectiveness of the RBF neural network approach     

基于改进的递归神经网络的网络安全态势预测

本文利用神经网络处理非线性、复杂性等优势,基于改进的递归神经网络预测网络安全态势,实验结果证明该方法运行效率较高,运行结果与实际值相比,误差较低,精确性较高。     

基于改进的递归神经网络的网络安全态势预测

本文利用神经网络处理非线性、复杂性等优势,基于改进的递归神经网络预测网络安全态势,实验结果证明该方法运行效率较高,运行结果与实际值相比,误差较低,精确性较高。     

Genetic algorithm based inversion of neural networks applied to optimised design of UWB planar antennas

An inversion of artificial neural networks using a genetic algorithm is presented for a novel concept of optimisation applied to UWB planar antennas of bow-tie type with respect to specified values of antenna performance in the frequency range 3.1-10.6 GHz. This efficient concept is shown to achieve significant reduction in computing time for optimisation. The multidimensional inversion is characterised by a simple composite fitness or target function that includes antenna parameters as a function of signal frequency or/and angle dependence. Good impedance matching and gain performance is achieved over the whole frequency range by adequately modifying the radiating contour profile of the conventional triangular bow-tie antenna.     

Indoor electric field level prediction model based on theartificial neural networks

A new field strength prediction model for the mobile phone environment is presented. The model is based on the principles of popular feedforward neural networks. Utilizing a new artificial neural network model some important disadvantages of both deterministic and empirical models can be overcome. In order to build the model, extensive field strength measurements were carried out inside two different buildings. The analysis of the model has shown that the proposed model is fast, accurate (on the order of the focal mean measurements uncertainty) and reliable     

Neural networks applied to traffic management in telephone networks

In this paper the application of neural networks to some of the network management tasks carried out in a regional Bell telephone company is described. Network managers monitor the telephone network for abnormal conditions and have the ability to place controls in the network to improve traffic flow. Conclusions are drawn regarding the utility and effectiveness of the neural networks in automating the network management tasks     

Macrocell electric field strength prediction model based upon artificial neural networks

A new macrocell prediction model for mobile radio environment is presented. The use of feedforward artificial neural networks makes it possible to overcome some important disadvantages of previous prediction models, including both deterministic and statistical types. Our sample implementation is based upon extensive electric field strength measurements (in the 900-MHz frequency band) that were carried out in the city of Belgrade using six different test transmitter locations. Comparison between the data obtained by the proposed electric field strength prediction model and independent measurement sets show that the proposed model is sufficiently accurate for use in planning mobile radio systems.     

Adaptive prediction applied to seismic event detection

Adaptive prediction was applied to the problem of detecting small seismic events in microseismic background noise. The Widrow-Hoff LMS adaptive filter [1], [2] used in a prediction configuration is compared with two standard seismic filters as an onset indicator. Examples demonstrate the technique's usefulness with both synthetic and actual seismic data.     

Optimal nonlinear adaptive prediction and modeling of MPEG video inATM networks using pipelined recurrent neural networks

This paper investigates the application of a pipelined recurrent neural network (PRNN) to the adaptive traffic prediction of MPEG video signal via dynamic ATM networks. The traffic signal of each picture type (I, P, and B) of MPEG video is characterized by a general nonlinear autoregressive moving average (NARMA) process. Moreover, a minimum mean-squared error predictor based on the NARMA model is developed to provide the best prediction for the video traffic signal. However, the explicit functional expression of the best mean-squared error predictor is actually unknown. To tackle this difficulty, a PRNN that consists of a number of simpler small-scale recurrent neural network (RNN) modules with less computational complexity is conducted to introduce the best nonlinear approximation capability into the minimum mean-squared error predictor model in order to accurately predict the future behavior of MPEG video traffic in a relatively short time period based on adaptive learning for each module from previous measurement data, in order to provide faster and more accurate control action to avoid the effects of excessive load situation. Since those modules of PRNN can be performed simultaneously in a pipelined parallelism fashion, this would lead to a significant improvement in the total computational efficiency of PRNN. In order to further improve the convergence performance of the adaptive algorithm for PRNN, a learning-rate annealing schedule is proposed to accelerate the adaptive learning process. Another advantage of the PRNN-based predictor is its generalization from learning that is useful for learning a dynamic environment for MPEG video traffic prediction in ATM networks where observations may be incomplete, delayed, or partially available. The PRNN-based predictor presented in this paper is shown to be promising and practically feasible in obtaining the best adaptive prediction of real-time MPEG video traffic     

Artificial neural networks and statistical modeling for electronic stress prediction using thermal profiling

Electronic components are constantly under stress due to factors such as signal density, temperature, humidity, and high current and voltage. Relatively little research has emphasized stress-level prediction under voltage stress. The purpose of this paper was to develop an electronic thermal profile model for stress-level prediction utilizing neural network (NN) and statistical approaches, such as multivariate regression models. Electronic components were removed from boards, subjected to different levels of stress, then replaced. An infrared camera was then used to capture information about component temperature changes over time under normal operating and stress conditions. Statistical analysis of the captured images suggests a strong correlation between thermal profiles and voltage stress levels. Artificial neural network (ANN) and statistical approaches were used to model temperature change profiles for components that had been stressed at different levels, and their predictive ability was compared. Separate data sets were used for model development and model verification. ANN prediction rates were around 70%, compared to 30% for the statistical approach. Experiments were also conducted to evaluate the robustness of the ANN model to the presence of noise in the data. Results suggested that the ANN model was able to accommodate the presence of noise. Various backpropagation (BP) learning algorithms were also evaluated and yielded similar average error rates. A 3-2-1 ANN topology performed better than 3-3-1 or 3-2-2-1 topologies, perhaps because the 3-2-1 topology has a higher data sample to nodes ratio than the other topologies.     

Artificial neural networks

Examines the following questions associated with artificial neural networks: why people are interested in artificial neural networks; what artificial neural networks are, from the point of view of electronic circuits, and how they work; how they can be programmed and made to solve particular problems; and whether interesting problems can actually be put on such networks. The author then describes the current state of artificial neural network technology and the resulting challenges to people working on electronic devices