Classification of biodegradable materials using QSAR modelling with uncertainty estimation

Abstract

The ability to determine the biodegradability of chemicals without resorting to expensive tests is ecologically and economically desirable. Models based on quantitative structure-activity relations (QSAR) provide some promise in this direction. However, QSAR models in the literature rarely provide uncertainty estimates in more detail than aggregated statistics such as the sensitivity and specificity of the model’s predictions. Almost never is there a means of assessing the uncertainty in an individual prediction. Without an uncertainty estimate, it is impossible to assess the trustworthiness of any particular prediction, which leaves the model with a low utility for regulatory purposes. In the present work, a QSAR model with uncertainty estimates is used to predict biodegradability for a set of substances from a publicly available data set. Separation was performed using a partial least squares discriminant analysis model, and the uncertainty was estimated using bootstrapping. The uncertainty prediction allows for confidence intervals to be assigned to any of the model’s predictions, allowing for a more complete assessment of the model that would be possible through a traditional statistical analysis. The results presented here are broadly applicable to other areas of modeling as well, because the calculation of the uncertainty will clearly demonstrate where additional tests are needed.

Mine landslide susceptibility assessment using IVM, ANN and SVM models considering the contribution of affecting factors

“…Bootstrap is a popular statistical method that are suitable for small sample. It is a random resampling technique [28] that can expand small samples into large samples, and it will be helpful for calculating the classification probability of each point. Traditional statistical methods like ROC are usually used to assess the classification results right or not, but sometimes, a classification result is neither right nor wrong, it is uncertainty within a range.…”

Classification of Biodegradable Substances Using Balanced Random Trees and Boosted C5.0 Decision Trees

“…In [ 15 , 17 , 18 ], the authors employed the UCI biodegradation dataset donated by Mansouri et al in [ 11 ]. In [ 15 ], the authors reduced the 41 molecular descriptors using principal components analysis (PCA) into four components.…”

Uncertainty estimation and misclassification probability for classification models based on discriminant analysis and support vector machines

“…The uncertainty estimation was based on Bootstrap [23], a random sampling method with replacement that allows confidence intervals to be placed on the model predictions based on uncertainties of the original data [10]. The procedure for calculating uncertainties based on residual bootstrap was originally presented by de Almeida et al [11] where J KLM represents the uncertainty propagation of SVM using RBF kernel.…”

“…The probability that sample ' is class k=1, denoted g , , is equivalent to the probability that 7 W is lower than the threshold value that separates the classes, 7 Z$_h . g , is given by the cumulative distribution function for the normal distribution [10]:…”

“…Uncertainty estimation for calibration models is well known [8,9]. However, for classification techniques, uncertainty estimation is still a new topic, so far mainly explored for partial least squares discriminant analysis (PLS-DA) [10,11]. Herein, we propose an uncertainty estimation method based on bootstrap for calculation of misclassification probabilities in linear discriminant analysis (LDA), quadratic discriminant analysis (QDA) and SVM models applied to four different datasets, where the classification stability is also evaluated by adding white Gaussian noise to the spectral data.…”