Fig. 1: Multi-objective Bayesian optimization workflow. | npj Computational Materials

Fig. 1: Multi-objective Bayesian optimization workflow.

From: Bayesian blacksmithing: discovering thermomechanical properties and deformation mechanisms in high-entropy refractory alloys

Fig. 1

The multi-objective function consists of optimizing both (static) thermomechanical properties and target a specific deformation mechanism in MoNbTaTi alloys. The initial database of properties is based on molecular dynamics (MD) simulations on a handful of compositions. The multi-objective Bayesian optimization is based on the Thompson sampling efficient multi-objective optimization (TS-EMO) algorithm34.

Back to article page