Development of a LSSVM-GC model for estimating the electrical conductivity of ionic liquids

In this communication, an extensive set of 1077 experimental electrical conductivity data for 54 ionic liquids (ILs) was collected from 21 different literature sources. Using this dataset, a reliable least square support vector machine-group contribution (LSSVM-GC) model has been developed, which employs a total of 22 sub-structures in addition to the temperature to represent/predict the electrical conductivity of ILs. In order to distinguish the effects of the anion and cation on the electrical conductivity of ILs, 11 sub-structures related to the chemical structure of anions, and 11 sub-structures related to the chemical structure of cations were implemented. The proposed model produces a low average absolute relative deviation (AARD) of less than 3.3% taking into consideration all 1077 experimental data values.