RL vs Supervised and Unsupervised Learning

RL learns from reward feedback through interaction, unlike supervised labels or unsupervised structure discovery.