一种基于蒙特卡罗方法的近红外波长选择算法

针对近红外光谱数据的特点，分析了基于偏最小二乘法(PLS)回归系数的波长选择方法，指出了其存在的问题，提出了一种新的波长选择算法。将PLS回归系数归一化为对应波长被选择的概率，并进行蒙特卡罗(Monte-Carlo)计算，即对不同的随机波长组合建立一系列PLS模型，预测误差最小的模型所对应的波长组合将被选择。这个过程可以在前一次波长选择的基础上重复进行，从而形成迭代算法。采用三个近红外数据集对提出的算法进行了验证，同时与基于PLS的无信息波长剔除法(UVE-PLS)和遗传算法(GA)进行了比较分析。实验结果表明，该方法能有效地提高波长选择的准确性和稳定性，在选择的波长点个数、模型的复杂度与预测误差方面，达到甚至优于现有算法，具有实用价值。

Abstract

Based on the feature of the (NIR) spectra, this paper analyses the method of wavelength selection using the partial least squaresc (PLS) regression coefficients and points out the existing problems, then proposes a new method for selecting wavelengths. It normalizes the PLS regression coefficients into the probability of the selected corresponding wavelengths, then a Monte-Carlo simulation based on the aforementioned probability is calculated. Some PLS models are constructed and evaluated using different random wavelengths combinations. The model with minimum predictive error is retained and the corresponding wavelength combinations are selected. This procedure can be iterated using the previous selected wavelengths to select fewer and fewer wavelengths. This method is tested on 3 NIR datasets and compared with the PLS-based uninformative variable elimination (UVE-PLS) and genetic algorithm (GA). Experimental results show that this method could select fewer wavelengths without sacrificing the complexity and predictive ability of the PLS model and could effectively improve the accuracy and stability of the wavelength selection.

参考文献

[1] 陆婉珍, 袁洪福, 徐广通等. 现代近红外光谱分析［M］. 北京: 中国石化出版社; 2000

Lu Wanzhen, Yuan Hongfu, Xu Guangtong et al.. Modern Analysis Techniques for Near Infrared Spectroscopy［M］. Beijing: China Petrochemical Press, 2000

[2] 聂黎行, 王钢力, 李志猛等. 近红外光谱法在中药生产过程分析中的应用［J］. 光学学报, 2009, 29(2): 541～547

Nie Lixing, Wang Gangli, Li Zhimeng et al.. Application of near infrared spectroscopy in process analysis of TCM manufacturing［J］. Acta Optica Sinica, 2009, 29(2): 541～547

[3] 张海红, 张淑娟, 王凤花等. 应用可见近红外光谱快速识别沙棘汁品牌［J］. 光学学报, 2010, 30(2): 574～578

Zhang Haihong, Zhang Shujuan, Wang Fenghua et al.. Study on fast discrimination of seabuckthorn juice varieties using visible-NIR spectroscopy［J］. Acta Optica Sinica, 2010, 30(2): 574～578

[4] 曹芳, 吴迪, 何勇等. 基于可见-近红外反射光谱技术的葡萄品种鉴别方法的研究［J］. 光学学报, 2009, 29(2): 537～540

Cao Fang, Wu Di, He Yong et al.. Variety discrimination of grapes based on visible-near reflection infrared spectroscopy［J］. Acta Optica Sinica, 2009, 29(2): 537～540

[5] Martens H, Naes T. Multivariate Calibration［M］. Chichester, UK: John Wiley & Sons, Inc., 1989

[6] 褚小立, 袁洪福, 陆婉珍. 近红外分析中光谱预处理及波长选择方法进展与应用［J］. 化学进展, 2004, 16(4): 528～542

Chu Xiaoli, Yuan Hongfu, Lu Wanzhen. Progress and application of spectral data pretreatment and wavelength selection methods in NIR analytical technique［J］. Progress in Chemistry, 2004, 16(4): 528～542

[7] J. A. F. Pierna, O. Abbas, V. Baeten et al.. A backward variable selection method for PLS regression (BVSPLS)［J］. Analytica Chimica Acta., 2009, 642(1～2): 89～93

[8] . Leardi, Application of genetic algorithm-PLS for feature selection in spectral data sets[J]. Journal of Chemometrics, 2000, 14: 643-655.

[9] 邹小波, 赵杰文. 用遗传算法快速提取近红外光谱特征区域和特征波长［J］. 光学学报, 2007, 27(7): 1316～1321

Zou Xiaobo, Zhao Jiewen. Methods of characteristic wavelength region and wavelength selection based on genetic algorithm［J］. Acta Optica Sinica, 2007, 27(7): 1316～1321

[10] 陈孝敬, 吴迪, 虞佳佳等. 一种用于可见-近红外光谱特征波长选择的新方法［J］. 光学学报, 2008, 28(11): 2153～2158

Chen Xiaojing, Wu Di, Yu Jiajia et al.. A new choice method of characteristic wavelength of visible/near infrared spectroscopy［J］. Acta Optica Sinica, 2008, 28(11): 2153～2158

[11] . J. Mitchell, J. J. Beauchamp. Bayesian variable selection in linear regression[J]. Journal of the American Statistical Association, 1988, 83(404): 1023-1032.

[12] . Bayesian linear regression and variable selection for spectroscopic calibration[J]. Analytica Chimica Acta, 2009, 631: 13-21.

[13] . G. Frenich, D. Jouan-Rimbaud, D. L. Massart et al.. Wavelength selection method for multicomponent spectrophotometric determinations using partial least squares[J]. Analyst., 1995, 120: 2787-2792.

[14] . F. Teófilo J. P. A. Martins, M. M. C., Ferreira MMC. Sorting variables by using informative vectors as a strategy for feature selection in multivariate regression[J]. Journal of Chemometrics, 2009, 23(1): 32-48.

[15] . Centner, D. L. Massart, O. E de Noord et al.. Elimination of uninformative variables for multivariate calibration[J]. Analytical Chemistry, 1996, 68(21): 3851-3858.

[16] . Stone, R. J. Brooks. Continuum regression: cross-validated sequentially constructed prediction embracing ordinary least squares, partial least squares and principal components regression[J]. Journal of the Royal Statistical Society. Series B(Methodological), 1990, 52(2): 237-269.

[17] H. A. Martens, P. Dardenne. Validation and verification of regression in small data sets［J］. Chemometrics and Intelligent Laboratory Systems, 1998, 44(1～2): 99～121

[18] . S. Xu, Y. Z. Liang. Monte Carlo cross validation[J]. Chemometrics and Intelligent Laboratory Systems, 2001, 56(1): 1-11.

[19] . D. R. Validation of regression models: methods and examples[J]. Technometrics, 1977, 19(4): 415-428.

[20] . H. Kalivas. Two data sets of near infrared spectra[J]. Chemometrics and Intelligent Laboratory Systems, 1997, 37(2): 255-259.

[21] Xiang Xianyi, Wen Zhiyu, Long Zaichuan et al.. Research on the near-infrared spectrometer system［J］. Spectroscopy and Spectral Analysis, 2009, 29(8): 2286～2290向贤毅, 温志渝, 龙再川等. 微型近红外光谱仪分析系统的研制［J］. 光谱学与光谱分析, 2009, 29(8): 2286～2290

洪明坚, 温泉, 温志渝. 一种基于蒙特卡罗方法的近红外波长选择算法[J]. 光学学报, 2010, 30(12): 3637. Hong Mingjian, Wen Quan, Wen Zhiyu. New Near Infrared Wavelength Selection Algorithm Based on Monte-Carlo Method[J]. Acta Optica Sinica, 2010, 30(12): 3637.

一种基于蒙特卡罗方法的近红外波长选择算法

关于本站 Cookie 的使用提示

全站搜索

一种基于蒙特卡罗方法的近红外波长选择算法

相关论文

相关资讯

关于本站 Cookie 的使用提示

全站搜索