Multi-Channel Subspace Mapping Using an Information Maximization Criterion

A new hybrid information maximization (HIM) algorithm is derived. This algorithm is able to perform subspace mapping of multi-channel signals, where the input (feature) vector for each of the channels is linearly transformed to an output vector. The algorithm is based on maximizing the mutual information (MI) between input and output sets for each of the channels, and between output sets across channels. Such formulation leads to a substantial redundancy reduction in the output sets, and the extraction of higher order features that exhibit coherence across time and/or space. In this paper, we develop the proposed algorithm and show that it combines efficiently the strengths of two well-known subspace mapping techniques, namely the principal component analysis (PCA) and the canonical correlation analysis (CCA). Unlike CCA, which is limited to two channels, the HIM algorithm can easily be extended to multiple channels. A number of simulations and real experiments are conducted to compare the performance of HIM to that of PCA and CCA.     

Time Series Segmentation Using a Novel Adaptive Eigendecomposition Algorithm

In this paper we present a new technique for time series segmentation built around a fast principal component analysis (PCA) algorithm that is on-line and stable. The traditional Generalized Likelihood Ratio Test (GLRT) has been used to solve the segmentation problem, but this has enormous limitations in terms of complexity and speed. Newer methods use gated experts and mixture models to detect transitions in time series. These techniques perform better than GLRT, but most of them require extensive training of relatively large neural networks. The segmentation method discussed in this paper is based on a novel idea that involves solving the generalized eigendecomposition of two consecutive windowed time series and can be formulated as a two-step PCA. Thus, the performance of our segmentation technique mainly depends on the efficiency of the PCA algorithm. Most of the existing techniques for PCA are based on gradient search procedures that are slow and they also suffer from convergence problems. The PCA algorithm presented in this paper is both online, and is proven to converge faster than the current methods.     

Computational Intelligence Prediction Model Integrating Empirical Mode Decomposition,Principal Component Analysis,and Weighted k-Nearest Neighbor

On the basis of machine leaning, suitable algorithms can make advanced time series analysis. This paper proposes a complex k-nearest neighbor (KNN) model for predicting financial time series. This model uses a complex feature extraction process integrating a forward rolling empirical mode decomposition (EMD) for financial time series signal analysis and principal component analysis (PCA) for the dimension reduction. The information-rich features are extracted then input to a weighted KNN classifier where the features are weighted with PCA loading. Finally, prediction is generated via regression on the selected nearest neighbors. The structure of the model as a whole is original. The test results on real historical data sets confirm the effectiveness of the models for predicting the Chinese stock index, an individual stock, and the EUR/USD exchange rate.     

Infrared Small Target Detection Using PPCA

Probabilistic PCA (PPCA) is an extension of PCA which reformulated PCA in a probabilistic framework. In this paper we propose a infrared small target detection algorithm using PPCA analogous to the face detection scheme using PCA, or known as “eigenface”. By computing the parameters of PPCA, we map the input vector from the image onto a subspace. After reconstructing the vector, the distance between the original vector and the reconstructed one will indicate the possibility of the input being a target. Experimental results show the effectiveness of this algorithm compared with other methods.     

Sparse Variable PCA Using Geodesic Steepest Descent

Principal component analysis (PCA) is a dimensionality reduction technique used in most fields of science and engineering. It aims to find linear combinations of the input variables that maximize variance. A problem with PCA is that it typically assigns nonzero loadings to all the variables, which in high dimensional problems can require a very large number of coefficients. But in many applications, the aim is to obtain a massive reduction in the number of coefficients. There are two very different types of sparse PCA problems: sparse loadings PCA (slPCA) which zeros out loadings (while generally keeping all of the variables) and sparse variable PCA which zeros out whole variables (typically leaving less than half of them). In this paper, we propose a new svPCA, which we call sparse variable noisy PCA (svnPCA). It is based on a statistical model, and this gives access to a range of modeling and inferential tools. Estimation is based on optimizing a novel penalized log-likelihood able to zero out whole variables rather than just some loadings. The estimation algorithm is based on the geodesic steepest descent algorithm. Finally, we develop a novel form of Bayesian information criterion (BIC) for tuning parameter selection. The svnPCA algorithm is applied to both simulated data and real functional magnetic resonance imaging (fMRI) data.      

利用脉冲耦合神经网络的高光谱多波段图像融合方法

针对高光谱图像波段众多、数据量大的特点,提出了一种基于脉冲耦合神经网络（Pulse Coupled Neural Networks, PCNN）模型的高光谱多波段图像融合方法.根据高光谱图像多输入的特点对原始PCNN模型进行了扩充,采用多通道PCNN模型来对输入图像进行非线性融合处理.通过分析传统变阈值衰减模型的特点及其不足,提出了修正的变阈值指数增加模型,以改善融合效果和降低PCNN运行的时间复杂度.利用记录点火时刻的赋时矩阵得到带有一定增强效果的融合结果图像.实验结果表明,该方法的融合效果要优于传统的主成分分析融合方法和小波变换融合方法.     

基于人工免疫RBF神经网络的时间序列预测方法研究

研究了一种基于免疫识别原理的径向基函数神经网络学习算法,该算法将所识别的数据作为抗原,抗体为抗原的压缩映射并作为神经网络模型的隐层中心,采用最小二乘法确定权值,提高了RBF神经网络收敛速度和精度.将人工免疫RBF神经网络应用于时间序列预测中,实例仿真结果证明了算法的有效性和可行性,为时间序列预测提供了一种新途径.     

基于LBP和深度学习的非限制条件下人脸识别算法

提出一种在非限制条件下,基于深度学习的人脸识别算法。同时,将LBP纹理特征作为深度网络的输入,通过逐层贪婪训练网络,获得良好的网络参数,并用训练好的网络对测试样本进行预测。在非限制条件下人脸库LFW上实验结果表明,该算法较传统算法（PCA、SVM、LBP）识别率高;另外,在Yale库和Yale-B库上也获得较高识别率,进一步说明以LBP纹理特征作为网络输入的深度学习方法能够对人脸图像进行准确识别。     

Hartley series and notch Hartley transform

A new class of Hartley transform is introduced—the Hartley series (HS). The Hartley series is appropriate in the situation that the input signal is continuous and periodic in time, and hence its Hartley transform is discrete in frequency. Through this class of Hartley transform, any continuous and periodic signal can be decomposed into the weighted sum of cas functions (where cas (·) = cos (·) + sin (·)) In order to compute these Hartley coefficients, an algorithm referred to as the notch Hartley transform (NHT) is also presented. This algorithm can be applied to the input signal composed of arbitrary frequencies, and all Hartley coefficients can be computed in advance of the end of one period of the signal     

基于数值微分的树突状细胞故障检测方法

针对现有树突状细胞算法（dendritic cell algorithm,DCA）在不同类型设备的故障检测中严重依赖人工经验定义输入信号,缺乏适应性和完备性,提出了一种基于数值微分的树突状细胞故障检测模型——NDDC-FD.首先,引入变化是系统危险发生的征兆和外在表现的思想,提出了一种基于变化危险感知的信号自适应提取方法,采用数值微分描述数据的变化,再从变化中提取输入信号.其次,原DC模型中异常抗原的评价方式对突变性故障能有效检测,却无法及时发现渐变性故障,提出了采用T细胞浓度作为故障评价指标.最后,在DAMADICS和TE两个基准平台上,将本文方法与原DCA算法和传统主元分析法（principal component analysis,PCA）进行比较测试.实验结果表明NDDC-FD方法不仅提高了原DCA算法的适应性,且和DCA、PCA相比具有较高检测率的同时,更能较早地检测到渐变性故障.因此,本文提出的故障检测方法NDDC-FD具有一般性,且性能良好.     

基于改进广义回归神经网络和主成分分析的宽带DOA估计

为了获取较高的宽带信号的DOA(direction-of-arrival)估计精度,提出了基于改进的广义回归神经网络(IGRNN,improved generalized regression neural network)和主成分分析(PCA,principalcomponent analysis)的宽带DOA估计算法。选用PCA方法对训练样本进行降维,以降低神经网络的复杂度;利用粒子群算法优化GRNN的参数;根据选取不同的聚焦角度确定粗估计、精估计的训练模型,通过粗估计得出目标的大致方位后,利用精估计模型得出最终的估计结果,避免了聚焦角度对估计精度的影响。仿真结果表明,本文提出的算法具有较好的估计精度和较高的工作效率。     

Gabor-based kernel PCA with doubly nonlinear mapping for face recognition with a single face image.

In this paper, a novel Gabor-based kernel principal component analysis (PCA) with doubly nonlinear mapping is proposed for human face recognition. In our approach, the Gabor wavelets are used to extract facial features, then a doubly nonlinear mapping kernel PCA (DKPCA) is proposed to perform feature transformation and face recognition. The conventional kernel PCA nonlinearly maps an input image into a high-dimensional feature space in order to make the mapped features linearly separable. However, this method does not consider the structural characteristics of the face images, and it is difficult to determine which nonlinear mapping is more effective for face recognition. In this paper, a new method of nonlinear mapping, which is performed in the original feature space, is defined. The proposed nonlinear mapping not only considers the statistical property of the input features, but also adopts an eigenmask to emphasize those important facial feature points. Therefore, after this mapping, the transformed features have a higher discriminating power, and the relative importance of the features adapts to the spatial importance of the face images. This new nonlinear mapping is combined with the conventional kernel PCA to be called "doubly" nonlinear mapping kernel PCA. The proposed algorithm is evaluated based on the Yale database, the AR database, the ORL database and the YaleB database by using different face recognition methods such as PCA, Gabor wavelets plus PCA, and Gabor wavelets plus kernel PCA with fractional power polynomial models. Experiments show that consistent and promising results are obtained.     

改进特征空间的红外弱小目标背景建模法

为有效去除动态背景对弱小目标信号的干扰,提出改进特征空间的红外弱小目标背景建模法来抑制背景。先采用改进的各向异性滤波算法从空域角度进行滤波以约束图像各个组分的差异,紧接着取连续时间域上多帧滤波后的图像组成一个特征矩阵,借助于主成分分析法进行特征分解,最后将输入图像投影到特征空间上进行背景建模,同时为了适应动态变化的背景,在时域上以一定学习率来更新背景模型。实验结果表明,提出的算法比传统的算法取得更好的背景估计效果,结构相似性SSIM、对比度增益I和背景抑制因子BIF分别大于0.97、15.46和5.25。     

直接正交鉴别保局投影算法

针对保局投影(LPP)及其衍生出的算法在人脸识别时须先采用主成分分析(PCA)算法对高维样本降维后才能应用,本文基于正交鉴别保局投影(ODLPP,orthogonal discriminal locality pre-serving projection)算法,提出了一种直接ODLPP(DODLPP)算法,利用拉普拉斯矩阵性质进行了相应的矩阵分解,可直接从高维样本的原始空间中提取投影矩阵。为解决ODLPP算法的小样本问题,给出先求解局部类内散度矩阵的零空间,然后再最大化类间散度矩阵的求解思路。人脸库上的实验结果表明所提算法的有效性。     

多项式非线性系统多频输入稳态响应的计算

针对工程实际中广泛存在并且有着十分重要应用的一大类非线性电路和系统,即非线性项为幂级数形式的非线性系统,本文称之为多项式非线性系统,提出了一种多频稳态响应的递归化计算方法,将这种非线性系统在多频输入下的稳态响应计算问题化为不断求解同一个线性系统在不同多频输入下的稳态响应,并且基于所构建的算法原理,采用目前广泛使用的Matlab语言编制了通用程序.大量算例表明,本文所提出的方法可以十分有效的用于计算这类系统的多频稳态响应.     

基于递归正交最小二乘的径向基函数网络人脸识别

本文提出了一种基于递归正交最小二乘的径向基函数(RBF)网络人脸识别算法,该算法首先使用主成分分析(PCA)提取输入图像特征,将提取的特征作为RBF网络的输入进行识别,在求取网络权值时采用递归正交最小二乘(ROLS)算法。实验表明,该算法能明显地缩短训练时间同时具有较高的识别率。     

An input-delay neural-network-based approach for piecewise ECG signal compression

We propose an input delay neural network (IDNN) based time series prediction algorithm for compressing electrocardiogram (ECG) signals. Our algorithm has been tested and successfully compared vis-à-vis other popular techniques for its compression efficiency and reconstruction capability.     

A new method of nonstationary time series analysis based oninhomogeneous AR equation

We propose a new model for nonstationary time series analysis. The model is of a noise-contaminated signal of an AR system excited by a sequence of an input signal represented in terms of orthogonal functions. We also propose an algorithm that enables us to estimate parameters of the AR part and the input signal simultaneously. The models are finally evaluated by testing the recovery of an output signal. Examples of data analysis of the synthetic time series are shown for the ease in which the input signal is represented by a sequence of wavelets     

Fast principal component extraction by a weighted informationcriterion

Principal component analysis (PCA) is an essential technique in data compression and feature extraction, and there has been much interest in developing fast PICA algorithms. On the basis of the concepts of both weighted subspace and information maximization, this paper proposes a weighted information criterion (WINC) for searching the optimal solution of a linear neural network. We analytically show that the optimum weights globally asymptotically converge to the principal eigenvectors of a stationary vector stochastic process. We establish a dependent relation of choosing the weighting matrix on statistics of the input process through the analysis of stability of the equilibrium of the proposed criterion. Therefore, we are able to reveal the constraint on the choice of a weighting matrix. We develop two adaptive algorithms based on the WINC for extracting in parallel multiple principal components. Both algorithms are able to provide adaptive step size, which leads to a significant improvement in the learning performance. Furthermore, the recursive least squares (RLS) version of WINC algorithms has a low computational complexity O(Np), where N is the input vector dimension, and p is the number of desired principal components. In fact, the WINC algorithm corresponds to a three-layer linear neural network model capable of performing, in parallel, the extraction of multiple principal components. The WINC algorithm also generalizes some well-known PCA/PSA algorithms just by adjusting the corresponding parameters. Since the weighting matrix does not require an accurate value, it facilitates the system design of the WINC algorithm for practical applications. The accuracy and speed advantages of the WINC algorithm are verified through simulations