Pattern recognition for the classification of Raman spectroscopy signals
Guo Ping①; Lu Hanqing②; Du Weimin③
①College of Information Science Beijing Normal University Beijing 100875 China;②National Laboratory of Pattern Recognition CAS Beijing 100080 China;③Dept. of Physics Peking University Beijing 100871 China
Abstract：This paper presents a study on application of regularized Gaussian classifier to Raman spectroscopy signals pattern recognition. Each Raman spectrum is treated as a point in high dimensional vector space. For given samples, the statistical model is used and Bayes decision is adopted to classify then according to their maximum posterior probabilities. Computer experiments show that the classification accuracy is obtained as high as 99.81%.
2002, Vol. 24 
