Genetic Algorithms in Molecular Modeling 1996
DOI: 10.1016/b978-012213810-2/50008-6
|View full text |Cite
|
Sign up to set email alerts
|

Prediction of the Progesterone Receptor Binding of Steroids using a Combination of Genetic Algorithms and Neural Networks

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
3
1
1

Citation Types

1
17
0
4

Year Published

1996
1996
2003
2003

Publication Types

Select...
5

Relationship

0
5

Authors

Journals

citations
Cited by 7 publications
(22 citation statements)
references
References 26 publications
1
17
0
4
Order By: Relevance
“…The relative values of the Pearson's correlation coefficients corresponding to the training and the two test sets points towards the PNN with Gaussian kernel as the best balanced model. This comparison also suggests that the best models reported by van Helden et al [1] and So et al [2] are practically at par, and there are no sufficient performance reasons to claim superiority of either of these models over the other. Evaluating the idea that the PNN is overfitted, the overall poorer predictive performance of the two above quoted models may be interpreted as an indication that these two models are in fact subject themselves to more severe overfittingyovertraining than the PNN.…”
Section: Resultsmentioning
confidence: 68%
See 2 more Smart Citations
“…The relative values of the Pearson's correlation coefficients corresponding to the training and the two test sets points towards the PNN with Gaussian kernel as the best balanced model. This comparison also suggests that the best models reported by van Helden et al [1] and So et al [2] are practically at par, and there are no sufficient performance reasons to claim superiority of either of these models over the other. Evaluating the idea that the PNN is overfitted, the overall poorer predictive performance of the two above quoted models may be interpreted as an indication that these two models are in fact subject themselves to more severe overfittingyovertraining than the PNN.…”
Section: Resultsmentioning
confidence: 68%
“…Comparison of the Pearson's correlation coefficients will allow positioning the PNN-based model with respect to other reported models built on the same van Helden et al [1] data set.…”
Section: Resultsmentioning
confidence: 99%
See 1 more Smart Citation
“…Several applications can be found in the study of Quantitative Structure-Activity Relationships (QSAR). Dunn and Rogers [84] combine PLS with the model-generating ability of a GA to create genetic partial least squares, while a hybrid method combining GAs and neural networks was used by Van Helden et al [85]. Li et al [86] used non-linear PLS combined with a GA in a study of the fungicidal activity of a series of O-etil-N-isopropylphosphoro(thioureido)thioates, obtaining results better than those of the reference, while Hou et al [87] developed a QSAR program combining a GA with multiple linear regression and cross-validation.…”
Section: Application Of Gas In Regressionmentioning
confidence: 99%
“…Consequently, efficient models of the receptor binding of substances, and their agonistic and antagonistic effects are highly desirable. A comparatively small, but representative data set was developed and analyzed by Van Helden et al [23] for binding to the progesterone receptor. The same group [24] also used a combination of genetic algorithms and neural networks.…”
Section: D Qsars Of Vibrio Fischeri Ec50 Datamentioning
confidence: 99%