Title :
Evaluation of herbal medicine´s bioactivity capacity prediction algorithms
Author :
Hao Chen ; Poon, Josiah ; Poon, Simon ; Lizhi Cui ; Sze, Daniel
Author_Institution :
Univ. of Sydney, Sydney, NSW, Australia
Abstract :
Currently, there are many multivariate linear regression algorithms being used for predicting the bioactive capacity of herbal formulae or herbal extracts from their chromatographic fingerprints, such as Principal Component Regression (PCR), Partial Least Squares Regression (PLSR), Orthogonal Projections to Latent Structures (OPLS) and Elastic Net (EN). In this study, the predicting performance and the complexity of predictive models developed by PCR, PLSR, OPLS, and EN are evaluated and compared using a set of chromatographic fingerprints of Astragali Radix (AR) and their corresponding bioactivity: Cluster of Differentiation 80 (CD80).
Keywords :
bioinformatics; chromatography; least squares approximations; patient treatment; principal component analysis; regression analysis; OPLS; PCR; PLSR; astragali radix; chromatographic fingerprints; cluster-of-differentiation 80; elastic net; herbal extracts; herbal formulae; herbal medicine bioactivity capacity prediction algorithms; multivariate linear regression algorithms; orthogonal projections-to-latent structures; partial least squares regression; principal component regression; Australia; Biological system modeling; Complexity theory; Fingerprint recognition; Prediction algorithms; Predictive models; Training; bioactivity prediction; chromatographic fingerprint; evaluation;
Conference_Titel :
Bioinformatics and Biomedicine (BIBM), 2014 IEEE International Conference on
Conference_Location :
Belfast
DOI :
10.1109/BIBM.2014.6999313
