Evaluation of machine learning interpolation techniques for prediction of physical properties

A knowledge of the physical properties of materials as a function of temperature, composition, applied external stresses, etc. is an important consideration in materials and process design. For new systems, such properties may be unknown and hard to measure or estimate from numerical simulations such as molecular dynamics. Engineers rely on machine learning to employ existing data in order to predict properties for new systems. Several techniques are currently used for such purposes. These include neural network, polynomial interpolation and Gaussian processes as well as the more recent dynamic trees and scalable Gaussian processes. In this paper we compare these approaches for three sets of materials sciences data: molar volume, electrical conductivity and Martensite start temperature. We make recommendations depending on the nature of the data. We demonstrate that a thorough knowledge of the problem beforehand is critical in selecting the most successful machine learning technique. Our findings show that the Gaussian process regression technique gives very good predictions for all three sets of tested data. Typically, Gaussian process is very slow with a computational complexity of typically n(3) where n is the number of data points. In this paper, we found that the scalable Gaussian process approach was able to maintain the high accuracy of the predictions while improving speed considerably, make on-line learning possible. (C) 2014 Elsevier B.V. All rights reserved.