Machine learning and statistical analysis in material property prediction
MetadataShow full item record
With the development of algorithms, models and data-driven efforts in other areas, machine learning is beginning to make impacts in materials science and engineering. In this work, we review the basic steps of using machine learning in materials science. We also develop several machine learning methods to predict the two physically-distinct properties of transparent conductors: formation enthalpy, which is an indication of stability, and bandgap energy, which is an indication of optical transparency. These include regression-based models such as the ordinary least squares (OLS) regression model, stepwise selection model, Ridge model and Lasso model, and tree-based models such as the random forest model and gradient boosted model (GBM). We discuss the advantages and potential problems of each model and provide suggestions for possible applications.