Machine learning of molecular properties: locality and active learning

TL;DR

This paper introduces a local, active learning-based machine learning algorithm for molecular property prediction that achieves high accuracy with smaller datasets and better handles outliers, outperforming existing methods.

Contribution

The authors develop a novel local and active learning approach that improves accuracy and outlier handling in molecular property prediction with smaller training datasets.

Findings

High accuracy with small training sets

Significant reduction in outlier errors

Outperforms state-of-the-art algorithms on benchmarks

Abstract

In recent years the machine learning techniques have shown a great potential in various problems from a multitude of disciplines, including materials design and drug discovery. The high computational speed on the one hand and the accuracy comparable to that of DFT on another hand make machine learning algorithms efficient for high-throughput screening through chemical and configurational space. However, the machine learning algorithms available in the literature require large training datasets to reach the chemical accuracy and also show large errors for the so-called outliers - the out-of-sample molecules, not well-represented in the training set. In the present paper we propose a new machine learning algorithm for predicting molecular properties that addresses these two issues: it is based on a local model of interatomic interactions providing high accuracy when trained on relatively small training sets and an active learning algorithm of optimally choosing the training set that significantly reduces the errors for the outliers. We compare our model to the other state-of-the-art algorithms from the literature on the widely used benchmark tests.