Machine Learning Spectroscopy Based on Group Contribution and Molecule Contribution Methods

Abstract

A group contribution (GC) – molecule contribution (MC) – machine learning (ML) protocol for accurate prediction of absorption spectra is presented. Upon combination of ML with revised GC methods, both the maximum absorption wavelengths and the full spectra of various sorts of dyes are afforded accurately and efficiently – by using only a small data set for training. Further, by employing a MC method designed specifically for revGC and based on MC-interpretated mixing rule by, the spectra for mixtures are obtained with high accuracy. Considering that the revGC-MC-ML protocol combines chemical principles and data-driven tools, most likely, it will be proven efficient to solve molecular-property mattered problems in wider fields.