Fig. 3: Bayesian inference workflow. | npj Computational Materials

Fig. 3: Bayesian inference workflow.

From: Radiative properties of quantum emitters in boron nitride from excited state calculations and Bayesian analysis

Fig. 3

Candidate structures are initially generated through combinatorial enumeration of point defects in hBN and attributed a prior likelihood based on structural complexity. The likelihood of each defect is then updated using results from our DFT calculations. For the most promising defect structures, the likelihood is refined using GW-BSE calculations.

Back to article page