1992
DOI: 10.1002/cem.1180060605
|View full text |Cite
|
Sign up to set email alerts
|

Predictive ability of regression models. Part II: Selection of the best predictive PLS model

Abstract: SUMMARYA procedure called GOLPE is suggested in order to detect those variables which increase the predictivity of PLS models. The procedure is based on evaluating the predictive power of a number of PLS models built by different combinations of variables selected according to a factorial design strategy. Examples are given of the efficiency of this variable selection procedure, which shows how these predictive PLS models are better than those obtained by all variables and better than the corresponding ordinar… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

1
60
0
1

Year Published

1997
1997
2018
2018

Publication Types

Select...
6
3

Relationship

0
9

Authors

Journals

citations
Cited by 117 publications
(62 citation statements)
references
References 16 publications
(3 reference statements)
1
60
0
1
Order By: Relevance
“…Linear regression analysis of the independent and dependent variables was carried out using partial least squares (PLS) 26) regression. 3D-QSAR analysis was carried out in two steps.…”
Section: Comfa/comsia Models and Partial Least Squares Analysismentioning
confidence: 99%
“…Linear regression analysis of the independent and dependent variables was carried out using partial least squares (PLS) 26) regression. 3D-QSAR analysis was carried out in two steps.…”
Section: Comfa/comsia Models and Partial Least Squares Analysismentioning
confidence: 99%
“…Predicted pEC50 [30] values of compounds in both the training and testing sets were presented together with their actual pEC50 values in Table 2, and correlations between predicted and actual pEC50 in CoMFA model were presented in Figure 1. Overall, predicted EC50 values were very close to the corresponding actual values for compounds in both the training and testing sets.…”
Section: D-qsar Studymentioning
confidence: 99%
“…Then, 3D-QSAR analysis was carried out using the PLS technique. The cross-validation and the ONC were used to evaluate the performance of the models, ONC was determined with the highest cross-validated q 2 [30,31]. Then, the non-cross-validated correlation coefficient r 2 value, standard error of estimate (SEE), and F value and standard error were calculated according to the definitions in Sybyl 7.3 package, and as factors for estimating.…”
Section: Partial Least-squares Analysismentioning
confidence: 99%
“…Além das constantes de substituintes utilizadas em QSAR clássi-co 7,8,[29][30][31][32][33] , devem-se incluir na análise propriedades físico-quí-micas moleculares tais como área superficial e volume moleculares 34 , propriedades derivadas de cálculo de orbital molecular 9, 35-37 , variáveis indicadoras 7,8,38 , índices de similaridade [39][40][41][42] e índices topológicos 34 . A utilização de grandes conjuntos de dados em QSAR pressupõe a necessidade de algum tipo de método de seleção de variáveis, como por exemplo, a busca sistemática 43 , as redes neurais [44][45][46][47][48][49][50] , os algoritmos genéticos e evolucionários 43,45,[51][52][53][54][55][56] e os métodos multivariados 25,45,52,[57][58][59][60] . Estes métodos são utilizados para detectar combinações de variáveis capazes de fornecer equações de regressão com elevado coeficiente de correlação, baixo desvio-padrão ou elevado teste F, e que tenham algum potencial para tornarem-se modelos de QSAR.…”
Section: Seleção De Variáveis Independentesunclassified