NIRS feature extraction based on deep auto-encoder neural network |
| |
| Institution: | 1. College of Information Science and Engineering, China Ocean University, No. 238 Songling Road, Qingdao 266100, China;2. R&D Center, China Tobacco Yunnan Industrial Co., Ltd, No. 367 Hongjin Road, Kunming 650231, China;3. College of Information Science and Technology, Qingdao University of Science and Technology, No. 99 Songling Road, Qingdao 266061, China;1. FOSS Analytical A/S, Foss Allé 1, DK-3400 Hillerød, Denmark;2. Department of Food Science, University of Copenhagen, Rolighedsvej 30, DK-1958 Frederiksberg C, Denmark;3. Department of Applied Mathematics and Computer Science, Technical University of Denmark, Matematiktorvet, Building 321, DK-2800 Kgs. Lyngby, Denmark;1. Radboud University Nijmegen, Institute for Computing and Information Science, The Netherlands;2. Radboud University Nijmegen, Institute for Molecules and Materials, The Netherlands;3. Center for Mathematical Sciences, Merck, Sharp & Dohme, Oss, The Netherlands;1. Ningbo Institute of Technology, Zhejiang University, Ningbo 315100, PR China;2. College of Agriculture and Biotechnology, Zhejiang University, Hangzhou 310058, PR China;1. School of Information and Communication Engineering, North University of China, China;2. Engineering Technology Research Center of Shanxi Province for Opto-Electronic Information and Instrument, China;3. School of Science, North University of China, China |
| |
| Abstract: | As a secondary analysis method, Near Infrared Spectroscopy (NIRS) needs an effective feature extraction method to improve the model performance. Deep auto-encoder (DAE) can build up an adaptive multilayer encoder network to transform the high-dimensional data into a low-dimensional code with both linear and nonlinear feature combinations. To evaluate its capability, we experimented on the spectra data obtained from different categories of cigarette with the method of DAE, and compared with the principal component analysis (PCA). The results showed that the DAE can extract more nonlinear features to characterize cigarette quality. In addition, the DAE also got the linear distribution of cigarette quality by its nonlinear transformation of features. Finally, we employed k-Nearest Neighbor (kNN) to classify different categories of cigarette with the features extracted by the linear transformation methods as PCA and wavelet transform-principal component analysis (WT-PCA), and the nonlinear transformation methods as DAE and isometric mapping (ISOMAP). The results showed that the pattern recognition mode built on features extracted by DAE was provided with more validity. |
| |
| Keywords: | Feature extraction Near infrared spectroscopy Deep auto encoder Cigarette pattern recognition |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
