Open AccessESTIMASI HUBUNGAN KUANTITATIF STRUKTUR-AKTIVITAS (HKSA) MENGGUNAKAN ARTIFICIAL NEURAL NETWORKS (ANN)
Author(s) -
Supriyanto Supriyanto
Publication year - 2009
Publication title -
molekul
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.125
H-Index - 2
eISSN - 2503-0310
pISSN - 1907-9761
DOI - 10.20884/1.jm.2009.4.1.60
Subject(s) - artificial neural network , broyden–fletcher–goldfarb–shanno algorithm , quantitative structure–activity relationship , artificial intelligence , computer science , mathematics , biological system , pattern recognition (psychology) , statistics , machine learning , biology , computer network , asynchronous communication
The Quantitative structure-Activity Relationship (QSAR) study has been performed on pattern of structure-molecule relationship. Artificial Neural Network (ANN) model used to estimate pattern of enzyme activity structure-molecule and atomic location in three-dimension for compound of flavonoid as the predictors. Value of determination coefficient used to compare between actual value and value of estimating by ANN models based on enzyme’s wavelength, so resulting cross validating is obtained. We use Quasy-Newton algorithm with Broyden-Fletcher-Goldfarb-Shanno (BFGS) procedure.