结合KPCA和多尺度纹理的IKONOS遥感影像决策树分类

城市地物类型多样,空间分布复杂,具有很强的非线性特征。核主成分分析(KPCA)通过将特征空间映射到高维核空间,可以表达图像像素间的高阶关系,因而可以提取图像的非线性特征,同时提供一组相互独立的主成分。本文在加入多尺度纹理特征的基础上,以应用地物分布的空间细节信息;且利用核主成分分析(KP-CA)方法对光谱和纹理量提取非线性特征信息,增大类别之间的可分性;并结合决策树分类方法对IKONOS遥感影像分类。实验结果表明:KPCA能很好提取地物之间的非线性特征,结合KPCA和多尺度纹理的决策树分类方法能有效地提取地物类型,提取精度为79.3%,KAPPA系数为0.763.     

非线性特征提取和LSSVM在化工过程故障诊断中应用

提出利用非线性特征提取（核主成分分析（KPCA）和核独立成分分析）消除数据的不相关性,降低维数。核主成分分析利用核函数把输入数据映射到特征空间,进行线性主成分分析计算提取特征;核独立成分分析在KPCA白化空间进行线性独立成分分析（ICA）变换提取独立成分。提取的特征作为最小二乘支持向量机分类器的输入,构建融合非线性特征提取和最小二乘支持向量机的智能故障分类方法。研究了该方法应用到某石化企业润滑油生产过程的故障诊断中的有效性和可行性。     

基于KPCA和CPSO的故障检测方法

提出一种基于核主元分析(KPCA)和混沌粒子优化群(CPSO)算法的非线性故障检测方法。通过核函数完成非线性变换,将变量由非线性的输入空间转换到线性的特征空间来计算主元,构造平方预测误差统计量检测故障是否发生。为避免粒子群算法的早熟现象,利用混沌优化的搜索特性,将CPSO算法应用到KPCA核参数的优化中。变压器故障检测结果表明,与基于PCA、KPCA和 PSO-KPCA的故障检测方法相比,该方法的检测正确率较高。     

一种融合KPCA和KDA的人脸识别新方法

核判别分析（KDA）和核主成分分析（KPCA）分别是线性判别分析（LDA）和主成分分析（PCA）在核空间中的非线性推广,提出了一种融合KDA和KPCA的特征提取方法并应用于人脸识别中,该方法综合利用KDA和KPCA 的优点来提高人脸识别的性能。此外,还提出了一种广义最近特征线(GNFL)方法来构造有效的分类器。实验结果证明：提出的方法获得了更好的识别结果。     

基于核主分量分析的脸谱识别方法

此文提出了一种新的脸谱识别方法--基于核主分量分析(KPCA)的脸谱识别方法.首先利用KPCA方法提取脸谱图象的特征,然后利用线性支持向量机进行识别.KPCA的基本思想就是首先经过一个非线性映射,将输入空间的数据映射到一个高维的特征空间中,以求数据在特征空间中线性可分(或近似线性可分),然后在特征空间中进行标准的PCA提取主元,作为特征向量.同时,我们将脸谱识别的经典方法主分量分析(PCA)(特征脸方法)和最近提出的独立分量分析(ICA)脸谱识别方法与新方法进行了比较,并利用ORL脸谱库进行实验,实验结果显示,新的方法具有较高的识别率.     

基于规范化KDDA的人脸识别

传统的PCA和LDA算法受限于“小样本问题”,且对象素的高阶相关性不敏感。本文将核函数方法与规范化LDA相结合,将原图像空间通过非线性映射变换到高维特征空间,并借助于“核技巧”在新的空间中应用鉴别分析方法。通过对ORL人脸库的大量实验研究表明,本文方法在特征提取方面明显优于PCA,KPCA,LDA等其他传统的人脸识别方法,在简化分类器的同时,也可以获得高识别率。     

基于规范化KDDA的人脸识别算法

传统的PCA和LDA算法受限于“小样本问题”,且对像素的高阶相关性不敏感。论文将核函数方法与规范化LDA相结合,将原图像空间通过非线性映射变换到高维特征空间,并借助于“核技巧”在新的空间中应用鉴别分析方法。通过对ORL人脸库的大量实验表明,该方法在特征提取方面优于PCA,KPCA,LDA等其他方法,在简化分类器的同时,也可以获得高识别率。     

融合小波变换与KPCA的分块人脸特征抽取与识别算法

鉴于小波多尺度变换对高维图像特征具有良好的压缩和表达能力,提出了一种融合小波变换与KPCA(核主成分分析)方法的分块人脸特征抽取与识别算法。该算法首先对人脸图像进行分块小波变换,再根据图像块的位置分布选取不同的频率分量;然后对此分量进行KPCA特征抽取,并通过对抽取到的特征进行融合来得到最终人脸鉴别特征;最后利用支持向量机分类器进行特征分类与识别。通过对ORL和Yale标准人脸图像库的实验仿真结果表明,该算法不仅在识别性能和分类速度上明显高于传统的PCA方法及融合小波特征的KPCA方法,而且对于人脸光照、姿态和表情变化均具有良好的鲁棒性。     

特征有效提取的自适应核特征子空间方法

基于核的主成分分析(KPCA)方法能提取数据的非线性特征,但特征提取的效率却与训练样本集合的容量成反比.文中提出一种特征提取的自适应核特征子空间方法来快速有效地提取特征.该方法和KPCA方法在理论分析框架上是一致的,但通过自适应的选取核子空间的张成向量,能在提高特征提取效率的同时不影响特征提取的精度.针对模拟数据和MNIST数据的实验结果表明文中方法优于经典KPCA方法和参考方法.     

结合DCT与KPCA的人脸识别

核主成分分析是主成分分析在核空间中的非线性推广,能有效应用于人脸识别,但是识别过程时间开销过大仍是待解决的问题。提出了一种结合离散余弦变换和核主分量分析的人脸识别方法。对人脸图像进行离散余弦变换,选择部分系数重建图像,采用核主分量分析的方法提取人脸特征,采用最近邻分类器进行识别。在ORL人脸库上的仿真结果表明所提出的方法速度快,综合性能优于核主成分分析方法。     

基于最优保留遗传算法的非线性模糊预测控制

针对传统的模型预测控制不能很好解决具有严重非线性、不确定性的对象或过程的控制问题，提出将模糊模型用于描述对象的非线性动态特性，通过将模糊模型的输出反馈作为模型输入，从而构成了模糊多步预测器。采用一种收敛精度高、速度快的具有最优保留特性遗传算法(EGA)，依据模型预测输出在线滚动求解控制律的非线性预测控制算法。仿真结果表明该算法对一类非线性系统具有较快的响应速度和较强的抗干扰能力。     

基于最优保留遗传算法的非线性模糊预测控制

针对传统的模型预测控制不能很好解决具有严重非线性、不确定性的对象或过程的控制问题。提出将模糊模型用于描述对象的非线性动态特性。通过将模糊模型的输出反馈回来作为模型输入，从而构成了模糊多步预测器，采用一种收敛精度高、速度快的具有最优保留特性遗传算法(EGA)依据模型预测输出在线滚动求解控制律的非线性预测控制算法。仿真结果表明该算法对一类非线性系统具有较快的响应速度和较强的抗干扰能力。     

Predicting the Future

We present a novel methodology for predicting future outcomes that uses small numbers of individuals participating in an imperfect information market. By determining their risk attitudes and performing a nonlinear aggregation of their predictions, we are able to assess the probability of the future outcome of an uncertain event and compare it to both the objective probability of its occurrence and the performance of the market as a whole. Experiments show that this nonlinear aggregation mechanism vastly outperforms both the imperfect market and the best of the participants. We then extend the mechanism to prove robust in the presence of public information.     

Fold-flip bifurcation analysis on a class of discrete-time neural network

This paper reports fold-flip bifurcation on a class of discrete-time neural network. Sufficient conditions are given to demonstrate fold-flip bifurcation. By performing linear and nonlinear transformation, the normal form and versal unfolding are derived to obtain the bifurcation diagrams of the truncated normal form such as fold bifurcation, flip bifurcation, and the Neimark–Sacker bifurcation of the period-2 cycle. Some numerical simulations are given to support the analytic results.     

On the design of BSB neural associative memories using semidefinite programming

This article is concerned with the reliable search for optimally performing BSB (brain state in a box) neural associative memories given a set of prototype patterns to be stored as stable equilibrium points. By converting and/or modifying the nonlinear constraints of a known formulation for the synthesis of BSB-based associative memories into linear matrix inequalities, we recast the synthesis into semidefinite programming problems and solve them by recently developed interior point methods. The validity of this approach is illustrated by a design example.     

Sensor and actuator fault isolation by structured partial PCA with nonlinear extensions

Partial PCA based on principal component analysis (PCA) with ideas borrowed from parity relations is a useful method in fault isolation (J. Gertler, W. Li, Y. Huang, T.J. McAvoy, Isolation enhanced principal component analysis, AIChE Journal 45(2) (1999) 323–334). By performing PCA on subsets of variables, a set of structured residuals can be obtained in the same way as structured parity relations. The structured residuals are utilized in composing an isolation scheme for sensor and actuator faults, according to a properly designed incidence matrix. To overcome the limitations of PCA, nonlinear approaches based on generalized PCA (GPCA) and nonlinear PCA (NPCA) are proposed. The nonlinear methods are demonstrated on an artificial 2×2 system while simulation studies on the Tennessee Eastman process illustrate the linear method and some extensions.     

Mathematical foundations of frequency‐domain modeling of nonlinear circuits and systems using the arithmetic operator method

The arithmetic operator method (AOM) is a method for performing arithmetic operations on one or more signals that are described by their spectra. By extension, any analytic functional operation on the signals can be performed in the frequency domain using matrix‐vector operations. The mathematical foundation of AOM is presented as a numerically efficient convolution‐like procedure in this article. It is directly applicable to the behavioral modeling of nonlinear RF and microwave circuits and systems. © 2003 Wiley Periodicals, Inc. Int J RF and Microwave CAE 13, 473–495, 2003.     

Efficient nonlinear data-assimilation in geophysical fluid dynamics

New ways of combining observations with numerical models are discussed in which the size of the state space can be very large, and the model can be highly nonlinear. Also the observations of the system can be related to the model variables in highly nonlinear ways, making this data-assimilation (or inverse) problem highly nonlinear. First we discuss the connection between data-assimilation and inverse problems, including regularization. We explore the choice of proposal density in a Particle Filter and show how the ‘curse of dimensionality’ might be beaten. In the standard Particle Filter ensembles of model runs are propagated forward in time until observations are encountered, rendering it a pure Monte-Carlo method. In large-dimensional systems this is very inefficient and very large numbers of model runs are needed to solve the data-assimilation problem realistically. In our approach we steer all model runs towards the observations resulting in a much more efficient method. By further ‘ensuring almost equal weight’ we avoid performing model runs that are useless in the end. Results are shown for the 40 and 1000 dimensional Lorenz 1995 model.     

An automatic nonlinear correlation approach for processing of hyperspectral images

Hyperspectral imaging technology demands sophisticated processing techniques that offer precise characterizations of complex spectral signatures. A nonlinear correlator structure is implemented for interference mitigation and object recognition. A key asset is the correlator's applicability to both the spatial (two-dimensional) and spectral (one-dimensional) domains, thus ideal for hyperspectral processing. The process consists of a standard convolution summed with a nonlinear adaptive term. The premise is the same in each case but the mathematical implementation is different. By performing the correlation calculations in the frequency domain, the processing algorithm is efficient, robust, and well suited for implementation on a parallel processing computational architecture. The nonlinear correlator depends on two parameters and an algorithm to determine these parameters based only on the input image (two-dimensional) or spectral signature (one-dimensional) is presented. Based on the results with the selected spatial and spectral templates, a target is identified and the spatial coordinates as well as the spectral signature are input to the final fusion stage, which analyses both spectral and spatial signatures for a correct target identification. Several examples are given and insights to template (mask) selection are provided.     

Sensor fault diagnosis of nonlinear processes based on structured kernel principal component analysis

A new sensor fault diagnosis method based on structured kernel principal component analysis （KPCA） is proposed for nonlinear processes. By performing KPCA on subsets of variables, a set of structured residuals, i.e., scaled powers of KPCA, can be obtained in the same way as partial PCA. The structured residuals are utilized in composing an isolation scheme for sensor fault diagnosis, according to a properly designed incidence matrix. Sensor fault sensitivity and critical sensitivity are defined, based on which an incidence matrix optimization algorithm is proposed to improve the performance of the structured KPCA. The effectiveness of the proposed method is demonstrated on the simulated continuous stirred tank reactor （CSTR） process.