Gaussian Process Regression for Absorption Spectra Analysis of Molecular Dimers

TL;DR

This paper presents a machine learning approach using Gaussian Process Regression to efficiently determine molecular dimer parameters from absorption spectra, providing both accurate results and insights into parameter space.

Contribution

The study introduces GPR for spectroscopy parameter extraction, enabling rapid convergence and comprehensive parameter space analysis, which is novel in molecular spectra analysis.

Findings

GPR accurately retrieves interaction parameters of molecular dimers.

The method agrees with quantum chemical calculations.

It offers insights into parameter regions consistent with experimental spectra.

Abstract

A common task is the determination of system parameters from spectroscopy, where one compares the experimental spectrum with calculated spectra, that depend on the desired parameters. Here we discuss an approach based on a machine learning technique, where the parameters for the numerical calculations are chosen from Gaussian Process Regression (GPR). This approach does not only quickly converge to an optimal parameter set, but in addition provides information about the complete parameter space, which allows for example to identify extended parameter regions where numerical spectra are consistent with the experimental one. We consider as example dimers of organic molecules and aim at extracting in particular the interaction between the monomers, and their mutual orientation. We find that indeed the GPR gives reliable results which are in agreement with direct calculations of these parameters using quantum chemical methods.