Workflow material synthesis and design are proposed.
Desired functional material can be directly predicted on the basis of first principle calculations and machine learning.
The predicted material properties for common materials have good agreement with experimental material data.
Prediction of material combinations is achieved using the trained support vector machine.