基于独立成分分析的高光谱图像异常检测

针对高光谱图像中背景及目标先验知识未知条件下的异常目标检测问题,提出了一种基于独立成分分析(ICA)的异常探测算法.首先估计原始数据的虚拟维(VD)以确定要分离的独立成分个数,在此基础上进行快速独立成分分析(FastICA),然后基于平均局部奇异度选择含异常信息较多的独立成分,最后使用丰度量化算法得到异常目标的丰度图像...     

利用DSP底层结构提高MPEG-4编码的实时性

由于利用TMS320C6000系列的DSP底层结构可以提高图像压缩的实时性,本文在TMS320C6416平台上针对DSP底层结构对MPEG-4视频编码进行了优化,包括对8个运算单元能并行执行的特性进行汇编。针对DSP芯片两级高速缓存的工作特点,对编码器中代码和数据所占用的内存空间进行优化。利用EDMA级联特性在内存开辟双缓冲区,同时完成视频数据编码和传输的工作,从而有效地改善编码效率。实验结果表明：该编码器可以对512 pixel×512 pixel大小的灰度图像,以30 frame/s帧频进行压缩,实现了对视频图像的实时编码。     

Classical Molecular Dynamics Model for Coupled Two‐Component Plasmas – Ionization Balance and Time Considerations

The dynamic properties of ion‐electron two‐component plasmas (TCP) are studied by using classical molecular dynamics (MD) simulations. There is a variety of time dependent and structural results that MD is able to provide in complement to other methods, e.g., useful micro‐field sequences can be generated. The method deals with some specific difficulties: the mass ratio between ions and electrons enforces very small time‐steps appropriate to follow electrons motion while, ions must move significantly in order to build, self consistently, their spatial structure. This results in expensive simulations. Electron trajectories are trapped and de‐trapped with multiple electron collisions around ions resulting in the occurrence of quasi metastable bound electron states. An analysis of micro‐fields at neutral in a hydrogen plasma reveals the need to consider a complete hierarchy of time scales extended typically over 7 order of magnitude, i.e., from a time‐step: ～10^‐19s, to a time required to obtain statistical averages, ～10^‐11s. In order to extend the MD capabilities in representing real coupled plasmas a classical ionization/recombination process has been implemented allowing to follow the evolution of plasmas involving several ion stages and model the ionization balance. Here again TCP simulations deal with extended time‐scale providing information about relaxation of non equilibrium plasma states (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

Ion Heating in Dusty Plasma of Noble Gas Mixtures

Ion heating in dusty plasma of noble gas mixtures is studied by the observation of dust particles in stratified glow discharge. The particles and their formations can be used as a “contact‐free” probe of the ion flows. It is shown that under condition of experiments transition of dust particles into crystalline state in pure gases occur at much lower pressures in comparison to the case of gas mixtures. This observation is also supported by the evaluation of “effective” kinetic temperature of dust particles as defined from the velocity distribution function at the same set of discharge parameters. Absolute value of temperature of dust component in the mixture of helium and argon indicates important role of argon ionization process (© 2011 WILEY‐VCH Verlag GmbH & Co. KGaA, Weinheim)     

水面油污染物差分偏振光谱信号预处理算法研究

采用差分偏振光谱法在3～5μm波段对水面溢油污染物进行被动遥感时,所测差分偏振光谱是含有强大气吸收光谱信号与油污染物目标弱光谱信号的混合谱,这给油污染物光谱特征识别带来了困难。另外,受环境因素以及油膜自身张力影响,水面油膜厚度分布以及油膜表面粗糙度在测量过程中发生变化,从而使得连续测量的差分偏振光谱中油污染物光谱信息含量存在不同。利用这一特点,基于固定点迭代的快速主成分分析算法FastPCA设计了水面溢油污染物差分偏振光谱信号预处理算法。实验结果表明,该算法可以有效地将水面油污染物目标光谱特征信息从具有强大气吸收的混合差分偏振光谱信号中提取出来,通过光谱重构得到油污染物光谱特征信号,可用于进一步的定性、定量分析。     

Fixed-point algorithms for constrained ICA and their applications in fMRI data analysis

Constrained independent component analysis (CICA) eliminates the order ambiguity of standard ICA by incorporating prior information into the learning process to sort the components intrinsically. However, the original CICA (OCICA) and its variants depend on a learning rate, which is not easy to be tuned for various applications. To solve this problem, two learning-rate-free CICA algorithms were derived in this paper using the fixed-point learning concept. A complete stability analysis was provided for the proposed methods, which also made a correction to the stability analysis given to OCICA. Variations for adding constraints either to the components or to the associated time courses were derived too. Using synthetic data, the proposed methods yielded a better stability and a better source separation quality in terms of higher signal-to-noise-ratio and smaller performance index than OCICA. For the artificially generated brain activations, the new CICAs demonstrated a better sensitivity/specificity performance than standard univariate general linear model (GLM) and standard ICA. Original CICA showed a similar sensitivity/specificity gain but failed to converge for several times. Using functional magnetic resonance imaging (fMRI) data acquired with a well-characterized sensorimotor task, the proposed CICAs yielded better sensitivity than OCICA, standard ICA and GLM in all the target functional regions in terms of either higher t values or larger suprathreshold cluster extensions using the same significance threshold. In addition, they were more stable than OCICA and standard ICA for analyzing the sensorimotor fMRI data.     

基于信息显著性差异的绿色产业评价指标筛选模型及实证

依据广义的绿色产业内涵,通过信息显著性差异与信息冗余分析定量筛选指标,建立了包括绿色生产指标层、绿色生态环境指标层、绿色消费方式指标层的大连市绿色产业评价指标体系。本文的创新与特色一是通过负载系数大小差异进行信息显著性分析,删除负载系数小、对综合评价结果影响弱的指标。克服了现有研究先进行相关分析剔除冗余指标,再利用负载系数大小差异筛选出反映信息显著性大的指标的弊端。事实上,指标间较高的相关性是计算负载系数的必要前提。因此,利用负载系数筛选指标前是不可进行指标的相关分析降低指标间的相关性的。二是通过结合专家知识经验辨析指标的实际含义,仅对反映信息可能重复的指标使用相关分析,剔除同指标层内相关系数较大的两个指标中反映指标层信息能力相对较弱的指标。克服了仅仅通过相关系数不能准确衡量指标间信息重复程度的弊端。研究结果表明,最终建立的指标体系通过25．68％的指标反映了97．09％的原始信息。通过与同类的指标体系进行对比,对大连市绿色产业评价指标体系的合理性做了进一步的分析。     

Forecasting retained earnings of privately held companies with PCA and L1 regression

We use proprietary data collected by SVB Analytics, an affiliate of Silicon Valley Bank, to forecast the retained earnings of privately held companies. Combining methods of principal component analysis (PCA) and L¹/quantile regression, we build multivariate linear models that feature excellent in‐sample fit and strong out‐of‐sample predictive accuracy. The combined PCA and L¹ technique effectively deals with multicollinearity and non‐normality of the data, and also performs favorably when compared against a variety of other models. Additionally, we propose a variable ranking procedure that explains which variables from the current quarter are most predictive of the next quarter's retained earnings. We fit models to the top five variables identified by the ranking procedure and thereby, discover interpretable models with excellent out‐of‐sample performance. Copyright © 2013 John Wiley & Sons, Ltd.