Predicting properties of materials and phase transformation using theoretical and computational multi-scale methods involving artificial intelligence and machine learning (ML) is important and highly rewarding. We review the relevant methods and mention a few applications and examples. We demonstrate that theoretical guidance narrows the search space and accelerates materials discovery.

中文翻译：

---

加速材料发现的理论和计算方法

使用涉及人工智能和机器学习 (ML) 的理论和计算多尺度方法预测材料的特性和相变是重要且非常有益的。我们回顾了相关的方法，并提到了一些应用和例子。我们证明了理论指导缩小了搜索空间并加速了材料发现。