The complexity of a learning problem is often determined by the size and structure of the hypothesis space.
软件工程的目标是控制复杂性,而不是制造复杂性。
The most rewarding part of research is seeing your work translate into real-world impact.