Abstract
In this article, we describe training and validation of a machine learning model for the prediction of organic compound normal boiling points. Data are drawn from the experimental literature as captured in the NIST Thermodynamics Research Center (TRC) SOURCE Data Archival System. The machine learning model is based on a graph neural network approach, a methodology that has proven powerful when applied to a variety of chemical problems. Model input is extracted from a 2D sketch of the molecule, making the methodology suitable for rapid prediction of normal boiling points in a wide variety of scenarios. Our final model predicts normal boiling points within 6 K (corresponding to a mean absolute percent error of 1.32%) with sample standard deviation less than 8 K. Additionally, we found that our model robustly identifies errors in the input data set during the model training phase, thereby further motivating the utility of systematic data exploration approaches for data-related efforts.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.