基于模糊聚类的神经网络集成

针对差异性是集成学习的一个重要条件,研究基于模糊聚类技术提高神经网络集成差异性的方法。提取大量弱分类器的权值和阈值并作为模糊聚类的数据对象,然后将聚类结果作为集成网络中个体网络的权值和阈值,最后在标准数据集上进行仿真实验,证实方法的有效性。     

Face Orientation Normalization Using Eye Positions

Despite the fact that progress in face recognition algorithms over the last decades has been made, changing lighting conditions and different face orientation still remain as a challenging problem. A standard face recognition system identifies the person by comparing the input picture against pictures of all faces in a database and finding the best match. Usually face matching is carried out in two steps： during the first step detection of a face is done by finding exact position of it in a complex background （various lightning condition）, and in the second step face identification is performed using gathered databases. In reality detected faces can appear in different position and they can be rotated, so these disturbances reduce quality of the recognition algorithms dramatically. In this paper to increase the identification accuracy we propose original geometric normalization of the face, based on extracted facial feature position such as eyes. For the eyes localization lbllowing methods has been used： color based method, mean eye template and SVM （Support Vector Machine） technique. Experimental investigation has shown that the best results for eye center detection can be achieved using SVM technique. The recognition rate increases statistically by 28% using face orientation normalization based on the eyes position.     

基于改进的ReliefF算法的神经网络集成分类模型

为进一步提高集成学习中各个神经网络的差异性,该文采用了一种改进的特征选择方法一基于概率抽样的ReliefF算法,并将其引入到集成所用的Bagging方法中。实验结果表明,该文提出的基于改进的KelietT算法的神经网络集成分类模型的泛化能力优于Bagging方法。     

Using a Neural Network to Approximate an Ensemble of Classifiers

Several methods (e.g., Bagging, Boosting) of constructing and combining an ensemble of classifiers have recently been shown capable of improving accuracy of a class of commonly used classifiers (e.g., decision trees, neural networks). The accuracy gain achieved, however, is at the expense of a higher requirement for storage and computation. This storage and computation overhead can decrease the utility of these methods when applied to real-world situations. In this Letter, we propose a learning approach which allows a single neural network to approximate a given ensemble of classifiers. Experiments on a large number of real-world data sets show that this approach can substantially save storage and computation while still maintaining accuracy similar to that of the entire ensemble.     

基于改进的ReliefF算法的神经网络集成分类模型

为进一步提高集成学习中各个神经网络的差异性,该文采用了一种改进的特征选择方法-基于概率抽样的ReliefF算法,并将其引入到集成所用的Bagging方法中。实验结果表明,该文提出的基于改进的ReliefF算法的神经网络集成分类模型的泛化能力优于Bagging方法。     

Ensemble of Evolving Neural Networks in Classification

The ensemble of evolving neural networks, which employs neural networks and genetic algorithms, is developed for classification problems in data mining. This network meets data mining requirements such as smart architecture, user interaction, and performance. The evolving neural network has a smart architecture in that it is able to select inputs from the environment and controls its topology. A built-in objective function of the network offers user interaction for customized classification. The bagging technique, which uses a portion of the training set in multiple networks, is applied to the ensemble of evolving neural networks in order to improve classification performance. The ensemble of evolving neural networks is tested by various data sets and produces better performance than both classical neural networks and simple ensemble methods.     

Knowledge Extraction from Transducer Neural Networks

Previously neural networks have shown interesting performance results for tasks such as classification, but they still suffer from an insufficient focus on the structure of the knowledge represented therein. In this paper, we analyze various knowledge extraction techniques in detail and we develop new transducer extraction techniques for the interpretation of recurrent neural network learning. First, we provide an overview of different possibilities to express structured knowledge using neural networks. Then, we analyze a type of recurrent network rigorously, applying a broad range of different techniques. We argue that analysis techniques, such as weight analysis using Hinton diagrams, hierarchical cluster analysis, and principal component analysis may be useful for providing certain views on the underlying knowledge. However, we demonstrate that these techniques are too static and too low-level for interpreting recurrent network classifications. The contribution of this paper is a particularly broad analysis of knowledge extraction techniques. Furthermore, we propose dynamic learning analysis and transducer extraction as two new dynamic interpretation techniques. Dynamic learning analysis provides a better understanding of how the network learns, while transducer extraction provides a better understanding of what the network represents.     

Extraction of Logical Rules from Neural Networks

Three neural-based methods for extraction of logical rules from data are presented. These methods facilitate conversion of graded response neural networks into networks performing logical functions. MLP2LN method tries to convert a standard MLP into a network performing logical operations (LN). C-MLP2LN is a constructive algorithm creating such MLP networks. Logical interpretation is assured by adding constraints to the cost function, forcing the weights to ±1 or 0. Skeletal networks emerge ensuring that a minimal number of logical rules are found. In both methods rules covering many training examples are generated before more specific rules covering exceptions. The third method, FSM2LN, is based on the probability density estimation. Several examples of performance of these methods are presented.     

神经网络规则提取及其在特征带识别中的应用

基于神经网络的规则提取通常只是针对某些特定的数据集,缺乏领域专家的鉴别,实际应用效果难以判定。本文针对分子标记特征带识别这一实际问题,在准确率损失尽量小的情况下剪枝训练好的神经网络,利用同符号规则提取算法（SSRE）抽取规则,避免了繁琐的规则抽取过程。实验表明此方法的效果良好,结果得到生物专家的认可。     

基于粗糙集约简的神经网络集成及其遥感图像分类应用

为降低集成特征选择方法的计算复杂性,提出了一种基于粗糙集约简的神经网络集成分类方法。该方法首先通过结合遗传算法求约简和重采样技术的动态约简技术,获得稳定的、泛化能力较强的属性约简集;然后,基于不同约简设计BP网络作为待集成的基分类器,并依据选择性集成思想,通过一定的搜索策略,找到具有最佳泛化性能的集成网络;最后通过多数投票法实现神经网络集成分类。该方法在某地区Landsat 7波段遥感图像的分类实验中得到了验证,由于通过粗糙集约简,过滤掉了大量分类性能欠佳的特征子集,和传统的集成特征选择方法相比,该方法时间开销少,计算复杂性低,具有满意的分类性能。     

基于集成神经网络入侵检测系统的研究与实现

为解决传统入侵检测模型所存在的检测效率低,对未知的入侵行为检测困难等问题,对集成学习进行了研究与探讨,提出一种采用遗传算法的集成神经网络入侵检测模型,阐述了模型的工作原理和各模块的主要功能.模型通过遗传算法寻找那些经过训练后差异较大的神经网络进行集成.实验表明,集成神经网络与检测率最好的单个神经网络相比检测率有所提高.同时,该模型采用机器学习方法,可使系统能动态地适应环境,不仅对已知的入侵具有较好的识别能力,而且能识别未知的入侵行为,从而实现入侵检测的智能化.     

基于聚类遗传算法的神经网络规则抽取及应用

提出了一种基于Gabor滤波器和神经网络规则抽取的烘焙面包品质分类方法。滤波器对烘焙面包切片区域灰度图像直接进行小波变换,用能量均值"和均方差!来表示灰度图像的纹理特征,并基于对隐层神经元输出值聚类的遗传算法实现了对面包品质分类的规则抽取。实验结果表明了该方法的实用性和可行性。     

Robust Practical Point Stabilization of a Nonholonomic Mobile Robot Using Neural Networks

A control structure that makes possible the integration of a kinematiccontroller and a neural network (NN) computed-torque controller fornonholonomic mobile robots is presented. A combined kinematic/torque controllaw is developed and stability is guaranteed by Lyapunov theory. Thiscontrol algorithm is applied to the practical point stabilization problemi.e., stabilization to a small neighborhood of the origin. The NN controllercan deal with unmodeled bounded disturbances and/or unstructured unmodeleddynamics in the vehicle. On-line NN weight tuning algorithms that do notrequire off-line learning yet guarantee small tracking errors and boundedcontrol signals are utilized.     

基于Gabor方向特征及神经网络的车牌灰度字符图像识别

针对低分辨率灰度车牌号码数字识别问题,提出了一种利用网格技术和Gabor变换直接从灰度图像进行特征提取的新方法,并没计了一种集成型神经N-N模型来进行识别,对大量的实验数据进行识别实验得到99．26％的识别率,显示该方法是非常有效的。     

Evolving the Topology and the Weights of Neural Networks Using a Dual Representation

Evolutionary computation is a class of global search techniques based on the learning process of a population of potential solutions to a given problem, that has been successfully applied to a variety of problems. In this paper a new approach to the construction of neural networks based on evolutionary computation is presented. A linear chromosome combined to a graph representation of the network are used by genetic operators, which allow the evolution of the architecture and the weights simultaneously without the need of local weight optimization. This paper describes the approach, the operators and reports results of the application of this technique to several binary classification problems.     

用遗传BP网络进行图像边缘检测

该文提出了一种基于遗传算法与图像特征向量的边缘检测方法。由于噪声的干扰,常规的图像边缘检测方法往往效果不佳,因此在充分考虑边缘和噪声本质区别的基础上,构造具有较强抗噪能力的特征向量;然后用样本图像对多层前馈神经网络采用遗传学习算法和误差反向传播算法(BP)相结合进行训练,即先用遗传学习算法进行全局训练,再用BP算法进行精确训练,使网络收敛速度加快和避免局部极小。最后,将训练后的网络用于图像的边缘检测。实验证明这种方法是有效的。     

基于神经网络的机器人轨迹跟踪控制

针对机器人模型未知情况，讨论了用神经网络和反馈控制实现机械手的跟踪控制。提出一种基于参考误差的投影算法来训练网络权值，训练后网络输出能逼近期望的前馈力矩，并从理论上证明跟踪误差的收敛性。仿真结果表明方案具有较好的跟踪性能和较强的抗干扰能力。     

Automatic Classification of Biological Particles from Electron-microscopy Images Using Conventional and Genetic-algorithm Optimized Learning Vector Quantization

Automatic classification of transmission electron-microscopy images is an important step in the complex task of determining the structure of biologial macromolecules. The process of 3D reconstruction from a set of such images implies their previous classification into homogeneous image classes. In general, different classes may represent either distinct biochemical specimens or specimens from different directions of an otherwise homogenous specimen. In this paper, a neural network classification algorithm has been applied to a real-data case in which it was known a priori the existence of two differentiated views of the same specimen. Using two labeled sets as a reference, the parameters and architecture of the classifier were optimized using a genetic algorithm. The global automatic process of training and optimization is implemented using the previously described g-lvq (genetic learning vector quantization) [10] algorithm, and compared to a non-optimized version of the algorithm, Kohonen's lvq (learning vector quantization) [7]. Using a part of the sample as training set, the results presented here show an efficient (approximately 90%) average classification rate of unknown samples in two classes. Finally, the implication of this kind of automatic classification of algorithms in the determination of three dimensional structure of biological particles is discused. This paper extends the results already presented in [11], and also improves them.     

Control of a Robotic Wheelchair Using Recurrent Networks

This paper describes an adaptive neural control system for governing the movements of a robotic wheelchair. It presents a new model of recurrent neural network based on a RBF architecture and combining in its architecture local recurrence and synaptic connections with FIR filters. This model is used in two different control architectures to command the movements of a robotic wheelchair. The training equations and the stability conditions of the control system are obtained. Practical tests show that the results achieved using the proposed method are better than those obtained using PID controllers or other recurrent neural networks models