Support Vector Machine and Optimized Method for Spectral Analysis
LIN Ji-peng1, 2, LIU Jun-hua1
1. School of Electrical Engineering, Xi’an Jiaotong University, Xi’an 710049, China 2. School of Information Engineering, Chang’an University, Xi’an 710054, China
Abstract：According to support vector machine based on the regularization theory, a small scale machine study theory was proposed to solve the problem of multi-gas analysis, which is mainly restricted by the lack of experimental samples. With its well nonlinear mapping ability, the training error was decided to be zero and global optimal parameters were obtained, hence the cross-sensitivity of spectrum is preferably eliminated. In multi-component gas analysis，the results show that the cross-sensitivity decreased to 1/81. A method based on genetic algorithm and cross-validation was proposed to solve the parameters selection of support vector machine(SVM), which still lacks theory support. The optimal structure parameters were achieved by genetic random search algorithm, the mean square error(MSE)0.018 of the spectrometer was achieved in 20th generation, and MSE decreased by multi-times in the fore generations. This hints that the genetic algorithm SVM is more efficient and has better generalizing ability.
林继鹏1, 2,刘君华1. 光谱分析中的支持向量机方法及其性能优化[J]. 光谱学与光谱分析, 2006, 26(12): 2232-2235.
LIN Ji-peng1, 2, LIU Jun-hua1. Support Vector Machine and Optimized Method for Spectral Analysis. SPECTROSCOPY AND SPECTRAL ANALYSIS, 2006, 26(12): 2232-2235.
[1] ZHANG Yong-huai，LIN Ji-peng，LIU Jun-hua(张永怀，林继鹏，刘君华). Chinese Journal of Scientific Instrument(仪器仪表学报)，2004，26(4)：45. [2] ZHANG Yong-huai，LIU Jun-hua(张永怀，刘君华). Spectroscopy and Spectral Analysis(光谱学与光谱分析)，2004，24(3): 376. [3] Pontil M, Verri A. IEEE Trans. On Pattern Analysis and Machine Intelligence, 1998, 20: 637. [4] ZHANG Guang-jun, Lü Jun-fang, ZHOU Xiu-yin(张广军，吕俊芳，周秀银). Journal of Beijing University of Aeronautics and Astronautics(北京航空航天大学学报)，1996，22(6)：655. [5] ZHANG Guang-jun, Lü Jun-fang, ZHOU Xiu-yin(张广军，吕俊芳，周秀银). Chinese Journal of Scientific Instrument(仪器仪表学报)，1997，18(2)：134. [6] Hara H, Kishi N, Iwaoka H. Optical MEMS, 2000 IEEE/LEOS International Conference on, 2000. 139. [7] Burges C J C. Data Mining Knowl. Discovery, 1998, 22: 1. [8] Rojo-ALVAREZ J L, de Prado-Cumplido M, Anibal R. IEEE Trans on Signal Processing, 2004, 52: 155. [9] Childers W, Thompson E L Jr, Harris D B, et al. Atmospheric Environment, 2001, (35): 1923. [10] LIN Ji-peng, LIU Jun-hua(林继鹏，刘君华). Journal of Jilin University(吉林大学学报)，2005，23(4): 521. [11] LIN Ji-peng, LIU Jun-hua(林继鹏，刘君华). Journal of Xi’an Jiaotong Universtity(西安交通大学学报)，2005，38(6)：787. [12] LIN Ji-peng, LIU Jun-hua, LING Zhen-bao(林继鹏，刘君华，凌振宝). Journal of Jilin University(吉林大学学报)，2004，22(5): 453. [13] Zhang Y H, Zhou J L, Lin J P, et al. Group Technology & Production Modernization, 2003, 20(1): 51. [14] Amatoa F D, De Rosa M. Optics and Lasers in Engineering, 2002, 37: 533.
