Bond type restricted radial distribution functions for accurate machine learning prediction of atomization energies

TL;DR

This paper introduces a new kernel method combining atomic property weighted radial distribution functions with Gaussian kernels, improving the prediction of molecular atomization energies by capturing multidimensional molecular features more efficiently.

Contribution

It presents a closed-form analytic kernel based on AP-RDF descriptors that enhances machine learning predictions of atomization energies, integrating tensor products for better molecular representation.

Findings

Improved accuracy in atomization energy prediction.

More computationally efficient than Bag-of-Bonds.

Established connection between quantum similarity and machine learning kernels.

Abstract

Understanding the performance of machine learning algorithms is essential for designing more accurate and efficient statistical models. It is not always possible to unravel the reasoning of neural networks. Here we propose a method for calculating machine learning kernels in closed and analytic form by combining atomic property weighted radial distribution function (AP-RDF) descriptor with a Gaussian kernel. This allowed us to analyse and improve the performance of the Bag-of-Bonds descriptor, when the bond type restriction is included in AP-RDF. The improvement is achieved for the prediction of molecular atomization energies and is due to the incorporation of a tensor product into the kernel which captures the multidimensional representation of the AP-RDF. On the other hand, the numerical version of the AP-RDF is a constant size descriptor, and it is more computationally efficient than Bag-of-Bonds. We have also discussed a connection between molecular quantum similarity and machine learning kernels with first principles kind of descriptors.