# Evolutionary computing and machine learning for discovering of low-energy defect con fi gurations

  • What is the main idea of this paper? To discovering of low-energy defect configurations?

  • What is low-energy defect configuration? Point defects in materials. Finding the most stable defective structures remains a very chanllenging task.

  • Why they use machine learning? Finding the stable defective structures is hard, they put forward a method to explore the potential energy surface. The approach offers a systematic way for finding low-energy configurations of isolated point defects in solids.

  • What is objective of machine learning? To search for low-energy minima of the potential energy surface. The approach most commonly used to tackle this problem is based on domain knowledge: the most likely low-energy defect configuration are selected according to intuition and/or the results obtained on analogs systems.

  • How do you check the configuration is the most stable defect configuration? One of the main drawbacks commond to all EAs, namely the need of several evaluations of the fitness function before an optimal solution is found, which make their use in first-principles studies computationally very demonding.

  • The steps of their approach?

  1. Generating several structures before a converged solution. -> Use machine learning.
  • Does this kind of method can be replaced and why?