RL vs Supervised and Unsupervised

RL learns from reward feedback rather than labeled examples or unlabeled patterns.

How does reinforcement learning differ from supervised learning?