A New Feature Extraction Method of Near-Infrared Spectra Based on the Addition of Historical Data
LI Hao-guang1,2, LI Wei-jun1*, QIN Hong1, ZHANG Li-ping1, DONG Xiao-li1, YU Yun-hua2
1. Institute of Semiconductors, Chinese Academy of Sciences, Beijing 100083, China 2. College of Information and Control Engineering, China University of Petroleum, Dongying 257061, China
Abstract：In traditional qualitative analysis of near-infrared (NIR) spectra, the stability of recognition models is decreased when new varieties of samples are added into the model. In order to improve the robustness of the model, a new feature extraction method based on the addition of historical data was put forward. The NIR training samples will be collected first, after that the historical data of the same species is added to constitute a larger and richer dataset. Then, the pretreated data of these training samples is projected to the feature space, which is constructed by feature extraction using partial least squares (PLS) based on the above dataset. Subsequently, orthogonal linear discriminant analysis (OLDA) is employed to extract features of the projected data. 18 varieties of corn seeds were taken as study subject, the comparative experiments with and without historical data are implemented respectively, and then the biomimetic pattern recognition (BPR) method is applied to verify the efficiency of the method proposed. The results suggest that the method adopted can improve the robustness of recognition model more effectively compared with the method without historical data. It maintains the high correct recognition ratios when new varieties are added into the model. Besides that, the recognition effect on test sets of the different days remains the same basically in the condition of same PLS dimensions. Therefore, the dimension of feature extraction can be set to some fixed values in recognition software. In this way, it can keep out of the trouble of manually modifying the optimal PLS parameter in recognition software if new varieties need to be added into the model. The experiment results of the thesis manifested the effectiveness of the proposed method.
Key words：The near-infrared Spectra;Project;Qualitative analysis;Partial least square
李浩光1,2，李卫军1*，覃 鸿1，张丽萍1，董肖莉1，于云华2 . 一种添加历史数据的近红外光谱特征提取方法研究 [J]. 光谱学与光谱分析, 2016, 36(10): 3148-3153.
LI Hao-guang1,2, LI Wei-jun1*, QIN Hong1, ZHANG Li-ping1, DONG Xiao-li1, YU Yun-hua2 . A New Feature Extraction Method of Near-Infrared Spectra Based on the Addition of Historical Data . SPECTROSCOPY AND SPECTRAL ANALYSIS, 2016, 36(10): 3148-3153.
[1] YAN Yan-lu(严衍禄). Modern Instrumental Analysis·3rd ed.(现代仪器分析·第3版). Beijing: China Agricultural University Press(北京：中国农业大学出版社)，2010. [2] LU Wan -zhen, YUAN Hong-fu, XU Guang-tong, et al( 陆婉珍, 袁洪福, 徐广通, 等) . Modern Near Infrared Spectroscopy Analytical Technology·2nd ed.(现代近红外光谱分析技术·第2版). Beijing: China Petrochemical Press(北京: 中国石化出版社), 2007. [3] ZHU Er-yi，YANG Peng-yuan(朱尔一，杨芃原). Chemometrics Technology and Application(化学计量学技术及应用). Beijing：Science Press(北京：科学出版社)，2001. [4] YAN Yan-lu, CHEN Bin, ZHU Da-zhou(严衍禄, 陈 斌，朱大洲). Near Infrared Spectroscopy Analytical—Principles, Technology and Application(近红外光谱分析的原理、技术与应用) . Beijing: China Light Industry Press(北京: 中国轻工业出版社), 2007. [5] CAO Wu, LI Wei-jun, WANG Ping, et al(曹 吾，李卫军，王 平，等). Spectroscopy and Spectral Analysis(光谱学与光谱分析)，2014, 34(6): 1. [6] WANG Shou-jue(王守觉). First Step to Multi-Dimensional Space Biomimetic Informatics(多维空间仿生信息学入门). Beijing: National Defense Industry Press(北京：国防工业出版社)，2008. [7] Ji Guoli, Huang Guangzao, Yang Zijiang, et al. Chemometrics and Intelligent Laboratory Systems, 2015, 144：56. [8] Bi Yiming, Chu Guohai, Wu Jizhong, et al. Chinese Journal of Analytical Chemistry, 2015, 43(7)：1086. [9] Duda R O，Hart P E，Stork D G. Pattern Classification(模式分类). Translated by LI Hong-dong，YAO Tian-xiang，et al(李宏东，姚天翔，等译). Beijing：China Machine Press(北京：机械工业出版社)，2003. [10] Chen Quansheng, Hui Zhe, Zhao Jiewen, et al LWT-Food Science and Technology, 2014, 57(2)：502. [11] Yang J, Jin Z, Yang J Y, et al. Pattern Recognition, 2004, 37(10): 2097. [12] Wang Shoujue. Biomimetic Pattern Recognition and Multi-Weight Neuron. Beijing: National Defense Industry Press，2012.
