A strategy to apply machine learning to small datasets in materials science

Abstract

There is growing interest in applying machine learning techniques in the research of materials science. However, although it is recognized that materials datasets are typically smaller and sometimes more diverse compared to other fields, the influence of availability of materials data on training machine learning models has not yet been studied, which prevents the possibility to establish accurate predictive rules using small materials datasets. Here we analyzed the fundamental interplay between the availability of materials data and the predictive capability of machine learning models. Instead of affecting the model precision directly, the effect of data size is mediated by the degree of freedom (DoF) of model, resulting in the phenomenon of association between precision and DoF. The appearance of precision-DoF association signals the issue of underfitting and is characterized by large bias of prediction, which consequently restricts the accurate prediction in unknown domains. We proposed to incorporate the crude estimation of property in the feature space to establish ML models using small sized materials data, which increases the accuracy of prediction without the cost of higher DoF. In three case studies of predicting the band gap of binary semiconductors, lattice thermal conductivity, and elastic properties of zeolites, the integration of crude estimation effectively boosted the predictive capability of machine learning models to state-of-art levels, demonstrating the generality of the proposed strategy to construct accurate machine learning models using small materials dataset.

Guided patchwork kriging to develop highly transferable thermal conductivity prediction models

“…For example, using the mean and standard deviation of elemental descriptors and BOP, the best model developed on 110 compounds for log-scaled κ l gives RMSE of 0.096 [6]. In another model developed on 93 compounds, the difference between experimental and predicted values of κ l has been reported to lie within a factor of 1.5-2 [52]. Using the descriptors related directly to the physics of κ l , a prediction model developed on 120 compounds gives RMSE of 0.21 [22].…”

A Critical Review of Machine Learning of Energy Materials

“…When both α 1 and α 2 are nonzero, the elastic net model is obtained. The ridge and LASSO regression models have found wide applications in materials ML to solve the small‐data problem, compute phonons using compressive sensing, perform feature selection, and avoid overfitting …”

Empirical modeling of dopability in diamond-like semiconductors

“…18,21 In the absence of an analytic model or high-throughput defect calculations, statistical learning from experimental or computational data can serve as an alternative to create empirical models and rules of thumb to make predictions of dopability in new compounds. Machine learning has proven successful in understanding and predicting energy and entropy, 29 potentials and forces, [30][31][32] structure, physical, and elastic properties, [33][34][35][36][37][38] bandgap, 34,39,40 and defects, 41 as well as enabling high-throughput screening and discovery, [42][43][44][45][46] and guiding experimental synthesis. 47,48 In order to properly model and interpret dopability, the construction of an empirical dataset for cross-validation is of vital importance.…”