The concept of probably approximately correct (PAC) learning has been foundational in computational learning theory.
可能近似正确(PAC)学习的概念在计算学习理论中具有基础性意义。
The most important thing in architecture is what cannot be seen.