Classification and Discrimination for Traditional Chinese Medicine Nature Based on OSC-OPLS/O2PLS-DA

Abstract

the research for Chinese herbs’ warm and cold natures classification is a significative thing for clinical. The paper put forward a new model classification and discrimination for Traditional Chinese medicine(TCM)' nature based on orthogonal signal correction-orthogonal partial least squares-discriminant analysis (OSC-OPLS/O2PLS-DA) after normalization. The first, data preprocessing and normalization for the metabolites sample space’s data, and the results data 135 multiply 839 dimension consist of three sections: warm nature’s normalization sample, cold nature’s normalization sample, blank group normalization sample; The second, OSC, dimension reduction and noise reduction for the metabolites sample space’s data; the third, OPLS/O2PLS-DA, Generate classification and discrimination Method for Traditional Chinese medicine(TCM)' nature. The model was proved to be feasible and effective after tested with 6 type’s warm nature’s herbs, 6 type’s cold nature’s herbs.