1 / 12

Online convex optimization Gradient descent without a gradient

Online convex optimization Gradient descent without a gradient. Abie Flaxman CMU Adam Tauman Kalai TTI Brendan McMahan CMU. Goal: find x f(x) ¸ max z 2 S f(z) – . High-dimensional. } . Standard convex optimization. Convex feasible set S ½ < d Concave function f : S ! <.

liam
Download Presentation

Online convex optimization Gradient descent without a gradient

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Online convex optimizationGradient descent without a gradient Abie Flaxman CMU Adam Tauman Kalai TTI Brendan McMahan CMU

  2. Goal: find x f(x) ¸ maxz2Sf(z) –  High-dimensional }  Standard convex optimization Convex feasible set S ½<d Concave function f : S !<

  3. Steepest ascent • Move in the direction of steepest ascent • Compute f’(x) (rf(x) in higher dimensions) • Works for convex optimization • (and many other problems) x1 x2 x3 x4

  4. Typical application • Toyota produces certain numbers of cars per month • Vector x 2<d (#Corollas, #Camrys, …) • Profit of Toyota is concave function of production vector • Maximize total (eq. average) profit PROBLEMS

  5. Problem definition and results • Sequence of unknown (concave) functions • On tth month, choose production xt2 S ½<d • Receive profit ft(xt) • Maximize avg. profit • No assumption on distribution of ft 8 z, z’ 2 S || z – z’ || · D ||rf(z)|| · G

  6. First try Zinkevich ’03: If we could only compute gradients… f4(x4) f3(x3) f2(x2) f4 PROFIT f1(x1) f3 f2 f1 x4 x3 x2 x1 #CAMRYS

  7. Idea: 1-point gradient estimate With probability ½, estimate = f(x + )/ With probability ½, estimate = –f(x – )/ PROFIT E[ estimate ] ¼ f’(x) x x- x+ #CAMRYS

  8. In expectation, gradient ascent on For online optimization, use Zinkevich’s analysis of online gradient ascent [Z03] Analysis PROFIT x- x+ #CAMRYS

  9. d-dimensional online algorithm • Choose u 2<d, ||u||=1 • Choose xt+1 = xt +  u ft(xt+u)/ • Repeat x3 x4 x1 x2 S

  10. Hidden complication… Dealing with steps outside set is difficult! S

  11. Hidden complication… Dealing with steps outside set is difficult! S’ …reshape into Isotropic position

  12. Related work & conclusions • Related work • Online convex gradient descent [Zinkevich03] • One point gradient estimates [Spall97,Granichin89] • Same exact problem [Kleinberg04] (different solution) • Conclusions • Can estimate gradient of a function from single evaluation (using randomness) • Adaptive adversary ft(xt)=ft(x1,x2,…,xt) • Useful for situations with a sequence of different functions, no gradient info, one evaluation per function, and others

More Related