# L22_dynrule - Lecture 22 Subgradient Methods Lecture 22...

This preview shows pages 1–7. Sign up to view the full content.

Lecture 22: Subgradient Methods April 11, 2007

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Lecture 22 Outline Methods with Dynamic Stepsize Rule Directional Derivatives and Subgradients -Subgradients and -Subdifferentals Convex Optimization 1
Lecture 22 Method with Dynamic Stepsize with Known f * General Assumption The function f ( x ) is convex and dom f = R n The set X is nonempty closed and convex Theorem Assume that General Assumption holds and X * is nonempty. Then, { x k } generated by the subgradient method with the stepsize α k = f ( x k ) - f * s k 2 converges to an optimal solution. Convex Optimization 2

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Lecture 22 Proof We use the Basic Iterate Relation : for any y X and any k 0 , x k +1 - y 2 x k - y 2 - 2 α k ( f ( x k ) - f ( y )) + α 2 k s k 2 Letting y = x * for any x * X * , we obtain x k +1 - x * 2 x k - x * 2 - 2 α k ( f ( x k ) - f * ) + α 2 k s k 2 = x k - x * 2 - α k s k 2 2 f ( x k ) - f * s k 2 - α k = x k - x * 2 - ( f ( x k ) - f * ) 2 s k 2 The sequence { x k } is bounded, thus subgradient norms s k are uniformly bounded by some scalar L . Therefore, for all k and any x * X * , x k +1 - x * 2 x k - x * 2 - ( f ( x k ) - f * ) 2 L 2 (1) Convex Optimization 3
Lecture 22 Eq. (1) implies that ( f ( x k ) - f * ) 2 L 2 x k - x * 2 - x k +1 - x * 2 , and k =0 ( f ( x k ) - f * ) 2 L 2 x 0 - x * 2 Hence f ( x k ) f * . Since { x k } ⊆ X is bounded, it has limit points that belong to X by closedness of X . By continuity of f and the fact f ( x k ) f * , it follows that all limit points are optimal. We now show that { x k } has a unique limit point. To arrive at a contradic- tion, suppose there are at least two: say ˜ x and ˆ x with ˜ x = ˆ x . Let { x m } M and { x k } K be subsequences converging to ˜ x and ˆ x , respectively. By Eq. (1), it follows that for all m, k with m k , x m - x * < x k - x * for all x * X * Let x * = ˆ x , and m ∈ M and k ∈ K . Then by taking limits, we obtain lim m →∞ m ∈M x m - ˆ x lim k →∞ k ∈K x k - ˆ x implying that ˜ x - ˆ x = 0 - a contradiction. Convex Optimization 4

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
Lecture 22 Convergence Rate The convergence rate of the method with dynamic stepsize using f * is linear at best For a function f with sharp minima , i.e., such that for some η > 0 f ( x ) - f * η dist ( x, X * ) for all x The rate is linear x k - ˜ x * c k x 0 - ˜ x * for all k 0 where ˜ x * X * is the limit point of { x k } and
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

### What students are saying

• As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

Kiran Temple University Fox School of Business ‘17, Course Hero Intern

• I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

Dana University of Pennsylvania ‘17, Course Hero Intern

• The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

Jill Tulane University ‘16, Course Hero Intern