改进的小波神经网络在桥梁损伤中的预测研究

提出基于BP算法的小波神经网络改进算法。仿真结果表明它避免了BP 神经网络结构设计的盲目性和局部最优等非线性优化问题,简化了训练,具有较强的函数学习能力和推广能力。该算法成功应用于桥梁损伤预测,具有广泛的应用前景。     

机器学习算法在ALOS影像分类中的应用研究

首先介绍了CART、C5.0和概率神经网络三种机器学习算法的原理,然后以覆盖湖北省公安县的ALOS影像为数据源,从整体精度、对训练样本大小和噪声的敏感性三个方面对它们进行了比较分析。结果显示C5.0算法分类的整体精度最高,达到83.59%。概率神经网络受训练样本大小和噪声的影响最低:在训练样本大小降为原样本数据量的40%时,其精度为78.52%;噪声占训练样本量的10%时,精度只下降了4.3%。通过分析可以看出,在训练样本量充足时,C5.0算法的分类精度最好,而在样本不足或者包含噪声的情况下,使用概率神经网络算法能比其他两种算法取得更好的分类效果。     

基于过程神经元网络的时间序列预测方法

传统人工神经网络时间序列预测方法难以表达时间序列中的时间累积效应。为此,提出一种基于过程神经元网络的时间序列预测方法。采用双链结构的量子粒子群对过程神经元网络进行训练,以Mackey-Glass混沌时间序列预测为例进行实验。仿真结果表明,该方法的均方误差比普通神经网络低一个数量级。     

一种基于RBF-HMM的网络入侵检测方法研究

本文提出了基于RBF-HMM模型的网络入侵检测方法,给出了该模掣的训练和识别方法.因为HMM模型的分类决策能力和对不确定信息的描述能力不理想,而人工神经网络对动志时间序列的建模能力尚不尽如人意,所以将RBF神经网络集成到HMM框架中,用RBF神经网络为HMM提供状态概牢输出.通过RBF神经网络的粗分类,克服了HMM的缺...     

模糊神经网络建模方法的研究

近年来神经网络在建模中得到了广泛地应用,但其学习过程需要大量的训练样本以保证其结果的正确性,在工业过程建模中,神经网络因可采集与训练样本数少,且信息不全等困难,难以建立一定正确度的模。针对这一问题,本文以Ｇａｕｓｓ函数为隶属度函数形式改进模糊聚类的Ｃ－平均法,提出了模糊ＣＧ－平均法,对一同组数据的聚类结果证明了此方法的有效性,模糊神经网络在化工中的研究尚处于初级阶段,本文将模糊ＣＧ－平均法与神经网络结合,构造由模糊化层、隶属度生成层、推理层及反模糊化输出层构成的模糊神经网络,实例表明本文所构造的模糊神经网络在使用较少训练样本的条件下仍能取得理想的结果,有助于直接从生产中建立所需的模型。     

前馈神经网络权值学习的复合形法

本文介绍一种新的前馈神经网络的随机学习方法，着重讨论该算法的实现，并讨论了将它和ＢＰ算法相结合，从而得到一种非常实用的神经网络学习算法。     

Improving Bayesian Regularization of ANN via Pre-training with Early-Stopping

We propose a simple method that enhances the performance of Bayesian Regularization of Artificial Neural Network (ANN) through pre-training of initial network with the Early-Stopping algorithm. The proposed method is applied to the regularization of Feed-forward Neural Networks to regress three benchmark data series. Significant reduction in both the cross-validation error and the number of training over standard Bayesian Regularisation is achieved.     

卷积神经网络应用于先心病心音信号分类研究

心脏听诊是先心病初诊和筛查的主要手段。传统心音分类算法普适性差，过程复杂，不利于将来实时化决策。采用1 800个心音信号对几种时间序列分类的主流深度学习网络进行训练，结果显示循环神经网络易出现过拟合；长短时记忆网络分类损失值0.257，准确率0.872；卷积神经网络损失值0.25，准确率0.896。实验表明卷积神经网络相比较其他两种网络具备更大的潜力。基于卷积神经网络的先心病分类算法，因训练样本量大，使网络普适性得到了保证。与其他分类器相比，CNN的另一个优势是其可自动提取特征。该研究有望用于机器辅助听诊。     

Generalized Regression Neural Networks and Feed Forward Neural Networks for prediction of scour depth around bridge piers

In this study, Generalized Regression Neural Networks (GRNN) and Feed Forward Neural Networks (FFNN) approaches are used to predict the scour depth around circular bridge piers. Hundred and sixty five data collected from various experimental studies, are used to predict equilibrium scour depth. The model consisting of the combination of dimensional data involving the input variables is constructed. The performance of the models in training and testing sets are compared with observations. Then, the model is also tested by Multiple Linear Regression (MLR) and empirical formula. The results of all approaches are compared in order to get more reliable comparison. The results indicated that GRNN can be applied successfully for prediction of scour depth around circular bridge piers.     

基于LVQ算法的SOM神经网络在入侵检测系统中的应用

目前,入侵检测技术（IDS）已成为网络安全领域研究的焦点,神经网络被应用到这项技术的研究上．文章在建立一、类基于SOM神经网络的分类器的基础上,应用了LVQ算法对SOM进行二次监督学习训练,极大提高了分类器的检测性能。仿真试验结果证明了该检测模型的有效性。     

遗传神经网络在模拟电路故障诊断中的应用

故障诊断对于事故后快速恢复具有重要的意义.模拟电路故障诊断有许多方法,提出了一种基于遗传算法优化的BP神经网络智能诊断技术.该方法采用基于实数编码的遗传算法优化神经网络权值和阈值,代替了原来BP网络随机设定的初始权值和阈值.然后再用改进的BP算法用已由遗传算法确定的空间对网络进行精确搜索.实验仿真结果表明基于遗传算法优化过的神经网络的训练步数得到大大的减少,泛化能力也得到提高.克服了传统BP算法的收敛速度慢,容易陷入局部极小的缺点.     

Novel Approaches to Probabilistic Neural Networks Through Bagging and Evolutionary Estimating of Prior Probabilities

In this contribution, novel approaches are proposed for the improvement of the performance of Probabilistic Neural Networks as well as the recently proposed Evolutionary Probabilistic Neural Networks. The Evolutionary Probabilistic Neural Network’s matrix of spread parameters is allowed to have different values in each class of neurons, resulting in a more flexible model that fits the data better and Particle Swarm Optimization is also employed for the estimation of the Probabilistic Neural Networks’s prior probabilities of each class. Moreover, the bagging technique is used to create an ensemble of Evolutionary Probabilistic Neural Networks in order to further improve the model’s performance. The above approaches have been applied to several well-known and widely used benchmark problems with promising results.      

基于粗糙集理论的模糊神经网络构造方法

构造模糊神经网络时确定初始的隶属函数是一个难点,提出了一种新的基于粗糙集理论的隶属函数获取算法,该算法根据粗糙集理论中基于属性重要性的离散化方法确定条件属性的断点,再通过断点确定各模糊集合隶属函数的中心和宽度,同时给出了网络各参数的修正公式;仿真结果证明,该算法在学习的快速性和精度上具有良好的性能.     

BP神经网络的置信度分析

随着神经网络在实际生产中日益广泛的应用，有必要对网络模型输出结果的精确度进行估计，本文介绍了一个种计算置信区间的方法，推导出目前广泛应用的ＢＰ网的置信区间计算公式。     

Multiprocessor scheduling and neural network training methods using shuffled frog-leaping algorithm

In this paper, we designed novel methods for Neural Network (NN) and Radial Basis function Neural Networks (RBFNN) training using Shuffled Frog-Leaping Algorithm (SFLA). This paper basically deals with the problem of multi-processor scheduling in a grid environment. We, in this paper, introduce three novel approaches for the task scheduling problem using a recently proposed Shuffled Frog-Leaping Algorithm (SFLA). In a first attempt, the scheduling problem is structured as a problem of optimization and solved by SFLA. Next, this paper makes use of SFLA trained Artificial Neural Network (ANN) and Radial Basis function Neural Networks (RBFNN) for the problem of task scheduling. Interestingly, the proposed methods yield better performance than contemporary algorithms as evidenced by simulation results.     

Artificial Neural Network trained by Particle Swarm Optimization for non-linear channel equalization

In this paper, we apply Artificial Neural Network (ANN) trained with Particle Swarm Optimization (PSO) for the problem of channel equalization. Existing applications of PSO to Artificial Neural Networks (ANN) training have only been used to find optimal weights of the network. Novelty in this paper is that it also takes care of appropriate network topology and transfer functions of the neuron. The PSO algorithm optimizes all the variables, and hence network weights and network parameters. Hence, this paper makes use of PSO to optimize the number of layers, input and hidden neurons, the type of transfer functions etc. This paper focuses on optimizing the weights, transfer function, and topology of an ANN constructed for channel equalization. Extensive simulations presented in this paper shows that, as compared to other ANN based equalizers as well as Neuro-fuzzy equalizers, the proposed equalizer performs better in all noise conditions.     

Advanced classification approach for neuronal phoneme recognition system based on efficient constructive training algorithm

This paper introduces a neural network optimization procedure allowing the generation of multilayer perceptron (MLP) network topologies with few connections, low complexity and high classification performance for phoneme’s recognition. An efficient constructive algorithm with incremental training using a new proposed Frame by Frame Neural Networks (FFNN) classification approach for automatic phoneme recognition is thus proposed. It is based on a novel recruiting hidden neuron’s procedure for a single hidden-layer. After an initializing phase started with initial small number of hidden neurons, this algorithm allows the Neural Networks (NNs) to adjust automatically its parameters during the training phase. The modular FFNN classification method is then constructed and tested to recognize 5 broad phonetic classes extracted from the TIMIT database. In order to take into account the speech variability related to the coarticulation effect, a Context Window of Three Successive Frame’s (CWTSF) analysis is applied. Although, an important reduction of the computational training time is observed, this technique penalized the overall Phone Recognition Rate (PRR) and increased the complexity of the recognition system. To alleviate these limitations, two feature dimensionality reduction techniques respectively based on Principal Component Analysis (PCA) and Self Organizing Maps (SOM) are investigated. It is observed an important improvement in the performance of the recognition system when the PCA technique is applied. Optimal neuronal phone recognition architecture is finally derived according to the following criteria: best PRR, minimum computational training time and complexity of the BPNN architecture.     

基于递归卷积神经网络的移动机器人定位算法

移动机器人定位已成为机器人研究的重要任务。提出基于递归卷积神经网络的移动机器人定位（Recurrent Convolutional Neural Networks-Based Mobile Robot Localization,RCNN-MRL）算法。递归卷积神经网络（Recurrent Convolutional Neural Networks,RCNN）结合卷积神经网络（Convolutional Neural Networks,CNN）和递归神经网络（Recurrent Neural Networks,RNN）的特性,并依据机器人上嵌入的照相机拍摄的第一人称视角图像,RCNN-MRL算法利用RCNN实现自主定位。具体而言,先通过RCNN有效地处理多个连续图像,再利用RCNN作为回归模型,进而估计机器人位置。同时,设计双轮机器人移动,获取多个时间序列图像信息。最后,依据双轮机器人随机移动建立仿真环境,分析机器人定位性能。实验数据表明,提出的RCNN模型能够实现自主定位。     

基于欧氏距离提高人工神经网的识别精度的方法

介绍了 Probabilistic Neural Networks(PNN)网的结构和算法 ,给出了点积和生长等相关概念的定义 ,研究了它们的性质 ,并用生长的方法改造训练样本集以便提高识别精度 .在模拟实验中 ,直接使用 PNN的算法 ,不能对水泥强度测试样本正确分类 ,使用本文提出的方法 ,能够对水泥强度的测试样本正确分类 ,识别率达到 10 0 % ,从而显示本文提出的方法是可行的和非常有效的     

进化神经网络研究综述

进化算法(EAs)与神经网络(NN)的结合已形成了一个新的领域一进化神经网络,在神经网络的研究中举足轻重。本文通过讨论和总结进化神经网络中的关键技术和现状,概述了其设计与构造的趋势。所讨论的是：(1)进化神经网络的研究方法;(2)进化模型;(3)应用实例及关键技术;(4)研究方向。