Premium
Investigation of Retention Behaviors of Essential Oils by Using QSRR
Author(s) -
Noorizadeh H.,
Farmany A.,
Khosravi Afra
Publication year - 2010
Publication title -
journal of the chinese chemical society
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.329
H-Index - 45
eISSN - 2192-6549
pISSN - 0009-4536
DOI - 10.1002/jccs.201000137
Subject(s) - chemistry , quantitative structure–activity relationship , test set , correlation coefficient , kovats retention index , artificial neural network , linear regression , biological system , kernel (algebra) , training set , partial least squares regression , molecular descriptor , artificial intelligence , genetic algorithm , pattern recognition (psychology) , chromatography , statistics , machine learning , mathematics , stereochemistry , computer science , combinatorics , gas chromatography , biology
Genetic algorithm and multiple linear regression (GA‐MLR), partial least square (GA‐PLS), kernel PLS (GA‐KPLS) and Levenberg‐Marquardt artificial neural network (L‐M ANN) technique were used to investigate the correlation between retention index (RI) and descriptors for diverse compounds in essential oils. The correlation coefficient cross validation (Q 2 ) between experimental and predicted retention index for training and test sets by GA‐MLR, GA‐PLS, GA‐KPLS and L‐M ANN was 0.948, 0.924, 0.958 and 0.980 (for training set), 0.917, 0.890, 0.915 and 0.954 (for test set), respectively. The L‐M ANN model with the final optimum network architecture of [5‐2‐1] gave a significantly better performance than the other models. This indicates that L‐M ANN can be used as an alternative modeling tool for quantitative structure‐property/retention relationship (QSPR/QSRR) studies.