水下噪声音色属性回归模型及其在目标识别中的应用

通过对声音的主观评价与客观分析而建立的主观感受数学模型,在许多领域都有重要的应用. 本文采用多元线性回归分析手段对水下噪声音色属性建立回归模型,提取音色特征并改善水下目标的识别效果. 首先,在前期水下噪声音色属性主观评价实验的基础上,将构成音色属性空间的5个成分的评价分值作为回归分析中的因变量,提取大量与听觉感知相关的听觉特征作为自变量;然后,通过相关分析和改进的逐步筛选法,挑选出反映音色属性的“最优”自变量子集;最后,利用向后剔除回归分析和水下目标识别实验,确定适当的音色模型,并通过假设检验证明该线性模型不仅正确有效,而且能改善水下目标识别效果.     

水下噪声音色属性回归模型及其在目标识别中的应用

通过对声音的主观评价与客观分析而建立的主观感受数学模型,在许多领域都有重要的应用. 本文采用多元线性回归分析手段对水下噪声音色属性建立回归模型,提取音色特征并改善水下目标的识别效果. 首先,在前期水下噪声音色属性主观评价实验的基础上,将构成音色属性空间的5个成分的评价分值作为回归分析中的因变量,提取大量与听觉感知相关的听觉特征作为自变量;然后,通过相关分析和改进的逐步筛选法,挑选出反映音色属性的“最优”自变量子集;最后,利用向后剔除回归分析和水下目标识别实验,确定适当的音色模型,并通过假设检验证明该线性模型不仅正确有效,而且能改善水下目标识别效果. 关键词： 音色 多元线性回归 主观评价     

基于小波核偏最小二乘回归方法的混沌系统建模研究

基于核学习的强大非线性映射能力,结合用于回归建模的线性偏最小二乘(PLS)算法,提出一种小波核偏最小二乘(WKPLS)回归方法. 该方法基于支持向量机使用的经典核函数技巧,将输入映射到高维非线性的特征空间,在特征空间中,构造线性的PLS回归模型. PLS方法利用输入与输出变量之间的协方差信息提取潜在特征,而可允许的小波核函数具有近似正交以及适用于信号局部分析的特性. 因此,结合它们优点的WKPLS方法显示了更好的非线性建模性能. 将WKPLS方法应用在非线性混沌动力系统建模上,并与基于高斯核的核偏最小二乘 关键词： 小波核 偏最小二乘回归 混沌系统 建模     

葡萄糖溶液近红外光谱分析中两种常用化学计量学方法的比较

采用主成分回归和偏最小二乘法分别对葡萄糖溶液近红外光谱分析 ,建立数学校正模型。主成分回归模型的相关系数为 0 .9780 ,偏最小二乘法模型相关系数为 0 .9986。对两种方法进行比较 ,确定对葡萄糖溶液近红外光谱分析的方法。     

基于可见-近红外光谱变量选择的荒漠土壤全磷含量估测研究

以新疆艾比湖湿地保护区采集的300个荒漠土壤样品为研究对象,利用ASD Field Spec○R 3 HR光谱仪获取的土壤可见-近红外光谱数据以及化学分析获取的土壤全磷数据为数据源,将原始光谱数据经过卷积平滑、标准正态变量变换以及一阶微分预处理后,采用蚁群-遗传结合区间偏最小二乘法提取荒漠土壤全磷含量特征波长,构建土壤全磷含量偏最小二乘回归预测模型;并与全谱偏最小二乘、蚁群-区间偏最小二乘、遗传-偏最小二乘模型进行比较。结果表明：经蚁群-区间偏最小二乘法筛选后,荒漠土壤全磷特征波段为500~700,1 101~1 300,1 501~1 700,1 901~2 100 nm;进一步采用遗传-区间偏最小二乘法进行变量选择,得到共线性最小的13个有效波长,分别为：1 621,546,1 259,573,1 572,1 527,564,1 186,1 988,1 541,2 024,1 118和1 191 nm。建模方法比较显示,采用蚁群-遗传结合区间偏最小二乘法选择的特征变量,建立的模型精度最高,其次是遗传算法、蚁群算法和全光谱。蚁群-遗传结合区间偏最小二乘法建立的土壤全磷含量的模型,效验证均方根误差RMSECV以及预测集均方根误差RMSEP分别为0.122和0.108 mg·g-1,效验证相关系数R_c以及预测集的相关系数R_p分别为0.535 7,0.555 9。因此,经过卷积平滑、标准正态变量变换以及一阶微分预处理,并利用蚁群-遗传结合区间偏最小二乘法建立的模型不仅简单,而且具有较高的预测精度和较好的稳健性,可以估算荒漠土壤全磷含量。     

基于高光谱的土壤有机质含量估算研究

高光谱遥感技术以其光谱分辨率高、波段连续性强、数据丰富的特点,因而在土壤养分研究中得到广泛应用.通过土壤钉机质的高光谱遥感分析,可以充分了解土壤养分的状况及动态变化,为指导农业生产及保护农业生态环境提供科学依据.本文基于江西省余江县和泰和县采集的34个红壤土样350～2 500 nm波段的光谱曲线,研究了土壤光谱与土壤有机质含量之间的关系.先对土壤反射率光谱进行两种变换:一阶微分(R')、倒数的对数log(1/R),然后在提取特征吸收波段的基础上,运用多元逐步线性回归法和偏最小二乘回归法建立相应的估算模型,并对模型进行检验.结果表明,偏最小二乘回归法优于多元逐步线性回归法,其建立的高光谱估算模型具有快速估算土壤中有机质含量的潜力.     

QSRR在化学和环境分析中应用研究进展

综述了QSRR中的多元线性回归（Multple Linear Regresion,MLR）、偏最小二乘回归（Partial Least Squares Regression,PLSR）、人工神经网络（Artificial Neural Networks,ANN）几种常用建模方法、分子描述符的选取的要点及原则及相关模型的建立和评价等方法在化学与环境化合物分析中的应用现状,并对QSRR在该领域的研究方向和发展前景进行了展望,以期为今后更深层次的研究作综合性的参考.     

用遗传算法快速提取近红外光谱特征区域和特征波长

提出了一种遗传区间偏最小二乘法(GA-iPLS),并用该方法快速提取苹果糖度近红外光谱的特征区域,在此基础上采用遗传偏最小二乘法(GA-PLS)提取苹果糖度近红外光谱的特征波长,进行苹果糖度预测。结果表明,整个光谱等分为40个子区间,遗传区间偏最小二乘法能快速寻找出5个特征子区间(第4,6,8,11,18号);在5个特征子区间的基础上用遗传偏最小二乘法继续优化,从中提取44个特征波长。建立在5个特征子区间和44个特征波长上的偏最小二乘法模型精度均优于全光谱偏最小二乘法模型,对预测集的预测相关系数提高了近10%;且模型得到了很大的简化,用于建模的主因子数减少了7个。这些结果表明,用这两种方法不但可以建立简洁、数据运算量少的模型,还可以快速地提取近红外光谱的特征区域和特征波长。     

堆叠监督自动编码器的近红外光谱建模

近红外光谱中包含了物质中有机分子含氢基团的特征信息,具有维度高、冗余大等特点.传统的基于浅层校正模型,比如主成分回归、偏最小二乘回归、人工神经网络、支持向量回归等,无法提取近红外光谱数据深层的信息.提出一种基于堆叠监督自动编码器的近红外光谱建模方法,不仅可以拟合光谱数据与理化值之间复杂的非线性关系,还可以提取数据深层的...     

基于光谱反射率的两种土壤有机质数学建模方法对比

已有土壤有机质的光谱预测模型其适用性受建模样本的采样尺度、土壤类型及光谱参数限制,需要在大尺度及范围上进一步检验适用性,并比较分析不同建模方法的建模效果以寻求适用性更好、精度更高的定量模型.在黑河上游大尺度范围采得225个土壤样品,进行了土壤有机质(SOC)及光谱反射率测定后将样本划分为建模集(180个土样)与验证集(45个土样).将土壤光谱反射率(R)变换处理后得到连续统去除(CR)、倒数(REC)、倒数之对数(LR)、一阶微分(FDR)及Kubelka-Munck变换系数共6种指标,针对建模集分别采用逐步线性回归与偏最小二乘回归方法建立12种光谱指标与SOC的数学模型,并采用验证集进行模型预测效果评价.结果表明:(1)采用逐步线性回归或偏最小二乘回归方法建模,LR指标对SOC变化的解释效果都是最好,是SOC的最优预测因子.(2)基于LR指标建立的SOC模型中,采用偏最小二乘回归模型比逐步线性回归模型的预测精度更好,相较于黑河上游已有的经验模型,偏最小二乘回归法建立的模型的预测效果也更好.(3)采用本实验的225个土壤样品对比验证了黑河上游仅有的SOC模型.该模型的SOC预测值与实测值通过了均值T检验且Pearson相关系数达0.826,表明在局部典型区域建立的SOC预测模型,可以应用到更大尺度上的土壤有机质预测研究.     

Quantum partial least squares regression algorithm for multiple correlation problem

Partial least squares (PLS) regression is an important linear regression method that efficiently addresses the multiple correlation problem by combining principal component analysis and multiple regression. In this paper, we present a quantum partial least squares (QPLS) regression algorithm. To solve the high time complexity of the PLS regression, we design a quantum eigenvector search method to speed up principal components and regression parameters construction. Meanwhile, we give a density matrix product method to avoid multiple access to quantum random access memory (QRAM) during building residual matrices. The time and space complexities of the QPLS regression are logarithmic in the independent variable dimension n, the dependent variable dimension w, and the number of variables m. This algorithm achieves exponential speed-ups over the PLS regression on n, m, and w. In addition, the QPLS regression inspires us to explore more potential quantum machine learning applications in future works.     

净信号预处理结合径向基偏最小二乘回归在血糖无创检测中的应用

为了提高人体血糖近红外光谱定量分析模型的预测精度,结合净信号预处理(NAP)算法和径向基偏最小二乘(RBFPLS)回归建立了一种适合于人体血糖测量的非线性建模方法NAP-RBFPLS。本文首先利用NAP对近红外光谱进行预处理来有效地提取原始光谱中仅与葡萄糖信号相关的光谱信息,从而有效地减弱了人体血液中水、白蛋白、血红蛋白、脂肪等成分的吸收干扰以及人体体温的变化、测量仪器本身的漂移、测量环境的变化和测量条件的变化引起的干扰因素与血糖变化的偶然相关问题;然后把净信号预处理后的近红外光谱数据通过RBFPLS建立了非线性定量分析模型来解决由于人体强散射引起的血糖浓度与近红外光谱之间的非线性关系,并与偏最小二乘(PLS)、基于净信号预处理的偏最小二乘(NAP-PLS)和RBFPLS这三种建模方法建立的定量分析模型进行了对比分析。实验结果表明,这两种方法相结合建立的非线性校正模型对预测集的预测精度有了很大的提高,这将对人体血糖浓度无创检测技术的研究具有实际应用价值。     

SG平滑和IBPLS联合优化水中油分析方法的研究

快速准确地检测水中矿物油的种类与含量对污染源及时排查和控制具有重要的意义,而红外光谱分析技术检测水中矿物油具有高效、快速、无污染的优势。为获得更加可靠的分析结果应用傅里叶变换衰减全反射红外光谱(FTIR-ATR)技术获取矿物油样品的光谱信息,采用SPXY法划分样本集。对偏最小二乘法(PLS)和迭代Bagging偏最小二乘法(IBPLS)这两种建立回归模型的方法进行对比分析,还比较了采用Savitzky-Golay(SG)平滑方法与迭代Bagging偏最小二乘法(IBPLS)相结合和单一采用迭代Bagging偏最小二乘法建立回归模型的区别。通过对预测回归曲线进行对比,得出通过SG平滑的预测效果明显优于未做的。而且采用SG平滑方法和IBPLS相结合的方法建立回归模型时,汽油模型参数RMSEP为0.001 125 g·mL-1,r为0.992 5;柴油模型参数RMSEP为0.001 384 g·mL-1 ,r为0.989 3。     

应用近红外高光谱成像技术预测甘蔗可溶性固形物含量

为了探究应用近红外高光谱成像技术对甘蔗内部可溶性固形物（SSC）预测的可行性,试验样本选择三种不同品种中的240个甘蔗节作为研究对象。通过高光谱成像系统获取甘蔗节的近红外光谱信息和图像信息,并分别探讨了光谱信息和图像纹理信息对甘蔗可溶性固形物预测的可行性。采用最小二乘回归（PLSR）,最小二乘支持向量机（LS-SVM）及主成分回归（PCR）建模方法构建甘蔗可溶性固形物的预测模型。比较了连续投影算法（SPA）、无信息变量消除算法（UVE）及区间偏最小二乘（iPLS）特征提取方法对预测结果的影响。实验结果表明：基于甘蔗的光谱信息能实现可溶性固形物的预测,其中偏最小二乘回归模型的建模集和预测集的相关系数分别为0.879和0.843,均方根误差分别为0.644和0.742。通过UVE算法提取105个有效波长所建立的PLSR模型的建模集及预测集相关系数分别为0.860和0.813,均方根误差分别为0.693和0.810。     

Optimized prediction of sugar content in ‘Snow’ pear using near-infrared diffuse reflectance spectroscopy combined with chemometrics

In near-infrared spectroscopy applications, the original spectra often contain redundant information, which will seriously affect the performance of chemometric models. Therefore, preprocessing, effective wavelengths selection, and appropriate regression models are essential. The objective of this study was to optimize the nondestructive determination multivariate calibration model of sugar content in ‘Snow’ pears, using near-infrared diffuse reflectance spectroscopy combined with chemometrics. All data (sugar content reference values and spectra data) from three measuring positions (P1, P2, and P3, marked around the pear’s equator at angular distances of approximately 120°) were divided into four grouped datasets, namely Set-1 (P1), Set-2 (P2), Set-3 (P3), and Set-4 (average of the three measuring positions). All subsequent optimized processes were performed based on each grouped dataset. First, different preprocessing methods were tested and an optimal method was determined. Then, synergy interval partial least squares and synergy interval partial least squares-competitive adaptive reweighted sampling were applied to select effective regions and effective wavelengths from all wavelengths, respectively, and partial least squares regression models were established and analyzed. In addition, support vector regression models were also established for comparative study. After comprehensive analysis of prediction accuracy and model complexity, the partial least squares regression model based on the 16 selected effective wavelengths for Set-4 was optimal, with the correlation coefficient for prediction, root-mean-square error of prediction, and residual predictive deviation of 0.9701, 0.2311, and 4.12, respectively. The results indicated that with these optimized processes, the multivariate calibration model of sugar content in ‘Snow’ pears was effectively optimized for each dataset. In addition, it is concluded that partial least squares regression was superior to support vector regression in this study, although some other researches had found different results in related fields.     

Influence of the Number of Loudspeakers on the Timbre in Horizontal and Mixed-Order Ambisoncis Reproduction

Ambisonics is a series of flexible spatial sound reproduction systems based on spatial harmonics decomposition of sound field. Traditional horizontal and spatial Ambisonics reconstruct horizontal and spatial sound field with certain order of spatial harmonics, respectively. Both the Shannon-Nyquist spatial sampling frequency limit for accurately reconstructing sound field and the complexity of system increase with the increasing order of Ambisonics. Based on the fact that the horizontal localization resolution of human hearing is higher than vertical resolution, mixed-order Ambisonics (MOA) reconstructs horizontal sound field with higher order spatial harmonics, while reconstructs vertical sound field with lower order spatial harmonics, and thereby reaches a compromise between the perceptual performance and the complexity of system. For a given order horizontal Ambisoncis or MOA reproduction, the number of horizontal loudspeakers is flexible, providing that it exceeds some low limit. By using Moore’s revised loudness model, the present work analyzes the influence of the number of horizontal loudspeakers on timbre both in horizontal Ambisonics and MOA reproduction. The binaural loudness level spectra (BLLS) of Ambisoncis reproduction are calculated and then compared with those of target sound field. The results indicate that below the Shannon-Nyquist limit of spatial sampling, increasing the number of horizontal loudspeakers influence little on BLLS then timbre. Above the limit, however, the BLLS for Ambisoncis reproduction deviate from those of target sound field. The extent of deviation depends on both the direction of target sound field and the number of loudspeakers. Increasing the number of horizontal loudspeakers may increase the change of BLLS then timbre in some cases, but reduce the change in some other cases. For MOA, the influence of the number of horizontal loudspeakers on BLLS and timbre reduces when virtual source departs from horizontal plane to the high or low elevation. The subjective evaluation experiment also validates the analysis.     

Selection of optimal combinations of inputs in a partial least squares model for prediction of soil organic matter

Abstract

Partial least squares model is widely used in estimation of soil physical and chemical parameters such as soil organic matter and moisture content, due to its advantages in dealing with collinearity of variables like hyperspectral reflectance. However, it is hard to determine optimal combination of partial least squares model input for soil organic matter prediction since there are lots of possibilities such as, different mathematical transformation of spectral reflectance, wavelength ranges, and spectral resolution. Laboratory hyperspectral reflectance of soils in Songnen plain were analyzed in this study, and the orthogonal experimental design method for deriving optimal combination of input variables for soil organic matter prediction models was introduced. For intercalating orthogonal experimental design table, five different levels which commonly used by researchers were assigned to factors. Results show that the optimal combination input for single black soil is using the derivative logarithmic reciprocal reflectance in the wavelength range selected by multiple stepwise regression at a spectral resolution of 5?nm (R²=?0.95, RMSE?=?0.21, and RPD?=?4.49), and different soils is using continuum removed in the wavelength range selected by MSR at a spectral resolution of 5?nm (R²?=?0.77, RMSE?=?0.74, and RPD?=?2.08). With optimal combination input, the partial least squares model prediction ability was evaluated as excellent for single black soil, possible for different soils. This study illustrates the orthogonal experimental design method can be an effective way to identify the optimal input variables of a partial least squares model for soil organic matter prediction, and multiple stepwise regression can be a preprocessing step to reduce hyperspectral data redundancy before using partial least squares to predict soil organic matter. Overall, this study provides a new approach for determining optimal input of partial least squares predicting model.     

基于高光谱的水体BOD含量模拟估算

高光谱技术由于满足连续性与光谱可分性的要求,具有能够区别同一种地物不同类别的能力,且光谱数据获取速度快,操作简易,在监测水体分布状况、水体指标上具有突出成就。生化需氧量BOD是评价水污染的重要指标,现行常规的测量方法为五日培养法,这种方法消耗试剂、操作复杂、受干扰因素多、测定时间长、不能及时反映水质变化、无法及时有效地预警突发水污染事件,鉴于传统方法的缺点,探索基于高光谱技术的水体BOD含量的估算和反演对水质评定具有重要意义。以西安地区三处地表水为研究区,共计60处试验点,每处试验点重复测定10次光谱与BOD值,取平均值作为原始光谱,利用Person相关系数法筛选光谱与BOD值的敏感波段,并使用主成分分析与最小二乘法消除光谱指标的多重共线性,建立水质BOD指标的多元线性回归模型与偏最小二乘回归模型。研究结果如下：（1）BOD敏感波段大体分布于600～900 nm,共筛选出了35个显著相关的原始光谱指标,其中758 nm相关系数绝对值最高（0.418）;（2）经由主成分分析降维得出的Z₁和Z₂与BOD指标的多元线性回归模型精度较好（R2=0.565,RMSE=0.007）,且主成分分析中可以明显区分0～0.2与0.4～0.6 mol·L-1 BOD浓度;（3）光谱指标与BOD指标构建偏最小二乘回归模型的精度R2高达0.896,RMSEP=0.746 9（留一交叉法均方根误差）;jack.test检验发现628 nm对反演水体BOD含量的影响极其显著,889与893 nm波段对其影响较为显著;(4）根据模型拟合精度,筛选的最优的BOD反演模型为偏最小二乘回归模型,对偏最小二乘模型进行精度检验,精度较好（R2=0.81）。基于以上试验结果,提出了一种基于偏最小二乘法高光谱水质BOD参数的反演方法,为水质BOD参数动态检测提供了新方法。