Descriptors in the Abraham equations for all the 209 PCBs have been obtained from literature chromatographic data. The six descriptors suffice for the prediction of water to octanol partitions, and for the prediction of various water to solvent and gas to solvent partitions, solubility in water, and water to micelle partition. For water to octanol, gas to dry octanol, gas to wet octanol, solubi...

The molecular weight and electrotopological E-state indices were used to estimate by Artificial Neural Networks aqueous solubility for a diverse set of 1291 organic compounds. The neural network with 33-4-1 neurons provided highly predictive results with r(2) = 0.91 and RMS = 0.62. The used parameters included several combinations of E-state indices with similar properties. The calculated resul...

A generic framework for enhancing an initial solubility prediction with ML, even simple methods and a modestly sized, sparse dataset. We dissect the setup to show model “locking on” target system as more data are made available.

The quantitative structure–property relationship (QSPR) method is used to develop the correlation between structures of crude oil hydrocarbons (80 compounds) and their boiling point and water solubility. Sub-structural molecular fragments (SMF) calculated from structure alone were used to represent molecular structures. A subset of the calculated fragments selected using stepwise regression (fo...

In this study fuzzy arithmetic is presented as a tool to tackle the prediction of the amount of barium, strontium and calcium sulfates scales in oilfield operations. Since the shape of fuzzy numbers’ membership functions is a spread representative of the whole possible values for a special model parameter, fuzzy numbers are able to consider the uncertainties in parameter determinations and thus...