The CatBoost as a tool to predict the isothermal compressibility of ionic liquids

In this work, we explore the possibility for calculating the isothermal compressibility of Ionic Liquids at atmospheric pressure by combination of the CatBoost machine learning library and Wada's rule. The CatBoost predicts the isothermal compressibility at 298.15 K using the density at the sam...

Full description

Saved in:
Bibliographic Details
Published in:Journal of molecular liquids 2021-07, Vol.333, p.115889, Article 115889
Main Authors: Postnikov, Eugene B., Jasiok, Bernadeta, Chorążewski, Mirosław
Format: Article
Language:English
Subjects:
Atmospheric pressure
CatBoost
Ionic liquids
Isothermal compressibility
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this work, we explore the possibility for calculating the isothermal compressibility of Ionic Liquids at atmospheric pressure by combination of the CatBoost machine learning library and Wada's rule. The CatBoost predicts the isothermal compressibility at 298.15 K using the density at the same temperature, the critical temperature and pressure, the molar mass, and the acentric factor. The obtained value plays the role of reference one combined with the power-law dependence on the density to get the target compressibility in a wider temperature range. The isothermal compressibilities for 35 Ionic Liquids, given in literature, were used for the comparison and its analysis. The average result at 298.15 K was 6.0% while for the whole range of temperatures was 6.2%. The influence of machine learning parameters on the prediction is discussed. •Combining numerical and categorical features for the prediction of isothermal compressibility coefficient.•Forecast's uncertainty quantification.•Quantitative analysis of quantities affecting the isothermal compressibility coefficient.
ISSN:0167-7322
DOI:10.1016/j.molliq.2021.115889