In practice, retraining a trained classifier is necessary when novel data become available. This paper adopts an incremental learning procedure to adaptively train a Kernel-based Nonlinear Representor (KNR), a recentl...In practice, retraining a trained classifier is necessary when novel data become available. This paper adopts an incremental learning procedure to adaptively train a Kernel-based Nonlinear Representor (KNR), a recently presented nonlinear classifier for optimal pattern representation, so that its generalization ability may be evaluated in time-variant situation and a sparser representation is obtained for computationally intensive tasks. The addressed techniques are applied to handwritten digit classification to illustrate the feasibility for pattern recognition.展开更多
Fourier transform infrared attenuated total reflectance （FTIR-ATR） spectroscopy was employed to characterize rapeseed oils. The spectral features of rapeseed oils were first investigated. Spectral data was processed...Fourier transform infrared attenuated total reflectance （FTIR-ATR） spectroscopy was employed to characterize rapeseed oils. The spectral features of rapeseed oils were first investigated. Spectral data was processed using principal component analysis （PCA） and linear discriminant analysis （LDA） to discriminate the oils from three cultivars of rapeseeds. As a result, 100% discrimination accuracy was obtained by LDA. Furthermore, the applicability of FTIR-ATR spectroscopy to characterize the changes of rapeseed oils caused by thermal treatment was studied. The rapeseed oil at 60 ℃ was regularly subjected to spectral measurement, and the spectral changes induced by thermal treatment were analyzed and discussed. This study had demonstrated the good performance of FTIR-ATR spectroscopy in characterizing rapeseed oils.展开更多
基金Supported by the Key Project of Chinese Ministry of Education (No.105150).
文摘In practice, retraining a trained classifier is necessary when novel data become available. This paper adopts an incremental learning procedure to adaptively train a Kernel-based Nonlinear Representor (KNR), a recently presented nonlinear classifier for optimal pattern representation, so that its generalization ability may be evaluated in time-variant situation and a sparser representation is obtained for computationally intensive tasks. The addressed techniques are applied to handwritten digit classification to illustrate the feasibility for pattern recognition.
文摘Fourier transform infrared attenuated total reflectance （FTIR-ATR） spectroscopy was employed to characterize rapeseed oils. The spectral features of rapeseed oils were first investigated. Spectral data was processed using principal component analysis （PCA） and linear discriminant analysis （LDA） to discriminate the oils from three cultivars of rapeseeds. As a result, 100% discrimination accuracy was obtained by LDA. Furthermore, the applicability of FTIR-ATR spectroscopy to characterize the changes of rapeseed oils caused by thermal treatment was studied. The rapeseed oil at 60 ℃ was regularly subjected to spectral measurement, and the spectral changes induced by thermal treatment were analyzed and discussed. This study had demonstrated the good performance of FTIR-ATR spectroscopy in characterizing rapeseed oils.
