Answers are grounded in this paper and cite their sources.
This study is an implementation of a robust jackknife-based descriptor selection procedure assisted with Gram-Schmidt orthogonalization. Selwood data including 31 molecules and 53 descriptors was considered in this study. Both multiple linear regression (MLR) and partial least squares (PLS) regression methods were applied during the jackknife procedures, and the desired results were obtained when using PLS regression on both autoscaled and orthogonalized data sets. Having used the Gram-Schmidt technique, descriptors were all orthogonalized, and their number was reduced to 30. A reproducible set of descriptors was obtained when PLS-jackknife was applied to the Gram-Schmidt orthogonalized data. The simple statistical t-test was applied to determine the significance of the obtained regression coefficients from jackknife resampling.Increasing the sample size, descriptors, based on their information content, were introduced into the model one by one and were sorted. The number of validated descriptors was in proportion with the sample size in the jackknife. The PLS-jackknife parameters, such as sample size and number and number of latent variables in PLS, and the starting descriptor in Gram-Schmidt orthogonalization were investigated and optimized.Applying PLS-jackknife to orthogonalized data in the optimized condition, five descriptors were validated with q²TOT2 ) 0.693 and R² ) 0.811. Compared to the previous reports, the obtained results are satisfactory.
No takes yet. Share an insight, caveat, or question.
Kompany‐Zareh et al. (2010) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: