Exercise · ZW-CONVERGENCE-WITH-LIPSCHITZ-GRADIENT-001 · ★☆☆ Basic

What does it mean for \( \nabla f \) to be Lipschitz with constant \( L \)?

Choose one option.

Show answer
  • A. That \( \|\nabla f(x)-\nabla f(y)\|\le L\|x-y\| \) for all \( x,y \)

Explanation

The Lipschitz definition bounds the difference of gradients by a constant times the distance between points. The option claiming \( f(x)\le L \) confuses a bound on the gradient with a bound on the function.

Study Continuous optimization in ZWMath →

Review the concept: Convergence with Lipschitz gradient

Result guaranteeing that gradient descent with a suitable step size converges for differentiable functions whose gradient \( \nabla f \) is Lipschitz continuous with constant \( L \).

\[f(x_k)-f^* \le \frac{L\|x_0-x^*\|^2}{2k}\]

More practice problems: convergence with lipschitz gradient

ZW-CONVERGENCE-WITH-LIPSCHITZ-GRADIENT-002 →