Multivariate Adaptive Regression Splines for Prediction of Rate Constants for Radical Degradation of Aromatic Pollutants in Water

Abstract

Listen

Translate article

A quantitative structure–activity relationship was developed to predict the rate constants for radical degradation of aromatic pollutants in water. A set of 1,508 zero-to three-dimensional descriptors was used for each molecule in the data set. Multiple linear regression was used as a descriptor selection method and the multivariate adaptive regression spline (MARS) method was successfully applied for the mapping model. The root-mean-square error and coefficient of determination were obtained as 0.0996 and 0.8998, respectively. In comparison with other models, the results show that MLR–MARS can be used as a powerful model for prediction of rate constants for radical degradation of aromatic pollutants in water.