Learning Rate
Domain: AI Fundamentals
🎯 Motivation
Controls how big each weight update step is. Too big diverges, too small is slow.
📋 Concrete Example
Learning rate 0.001 means each step changes weights by a tiny amount. 0.5 might overshoot and explode.
🔗 Analogy
Like adjusting a shower tap: tiny turns find the right temperature; big turns scald or freeze you.