Cooperative Model Framework with Minimal Viable Theoretical Data
Predicting material properties by solving the Kohn-Sham (KS) equation which is the basis of modern computational approaches to electronic structures has provided significant improvements in materials sciences. Despite its contributions, DFT or DFTB calculations are usually limited with the number of electrons and atoms in the studied systems. To overcome this difficulty, materials sciences communities have recently started adopting Machine Learning (ML) tools for a variety of reasons such as, reducing experimental cost, and time, designing more efficient systems, full automation, understanding complex systems etc. as an alternative to computations/simulations based on the KS equation. Here, we introduce a ML framework that can be used to accurately predict the KS total energy of anatase TiO2s nanoparticles (NPs) at different temperatures. Our process begins with generating the right data and end up with building the most optimal cooperative ML model. The proposed framework eliminates the need for experimental data and can be used over any NPs to figure out electronic structure and thus easily understand physical and chemical properties of NPs thanks to the ML model.