Correlation ranking and stepwise regression procedures in principal components artificial neural networks modeling with application to predict toxic activity and human serum albumin binding affinity |
| |
| Authors: | Omar Deeb |
| |
| Affiliation: | Faculty of Pharmacy, Al-Quds University, P.O. Box 20002, Jerusalem, Palestine |
| |
| Abstract: | |
| |
| Keywords: | ANN, artificial neural networks CR, correlation ranking CR-PC-ANN, correlation ranking principal components artificial neural networks CR-PC-ANN(C), correlation ranking principal components artificial neural networks based on the combined approach CR-PC-ANN(I), stepwise regression principal components artificial neural networks based on the individual approach FE, feature extraction FS, feature selection HOMO, highest occupied molecular orbital HSA, human serum albumin log IC50, logarithm of half maximal inhibitory concentration LMO-CV, leave-many-out cross validation log K'hsa, logarithm of HSA binding affinity LUMO, lowest unoccupied molecular orbital MLR, multiple linear regression PC, principal component PC-ANN, principal components artificial neural networks PCA, principal components analysis PCA(C), principal components analysis based on the combined approach PCA(I), principal components analysis based on the individual approach PCR, principal component regression R2, coefficient of determination R2CV, cross-validation coefficient of determination RMSE, root mean square error R2p, square of the correlation coefficient between the predicted and actual activities SR, stepwise regression SR-PC-ANN, stepwise regression principal components artificial neural networks SR-PC-ANN(C), stepwise regression principal components artificial neural networks based on the combined approach SR-PC-ANN(I), stepwise regression principal components artificial neural networks based on the individual approach SVD, singular value decomposition |
| 本文献已被 ScienceDirect 等数据库收录！ |
|
