What is 'Gradient Descent' convergence?

What is 'Gradient Descent' convergence?

Approach: Discuss the conditions and speed at which GD reaches a minimum.

Convergence depends on the learning rate and the curvature of the loss function. For convex functions, GD is guaranteed to reach the global minimum; for non-convex, it may settle in a local minimum.

Follow-up: What is the convergence rate of standard GD?

Key concepts: convexity, learning rate, lipschitz continuity.