SLIDE 1
Ben Cousins and Santosh Vempala Georgia Tech The Volume Problem - - PowerPoint PPT Presentation
Ben Cousins and Santosh Vempala Georgia Tech The Volume Problem - - PowerPoint PPT Presentation
Ben Cousins and Santosh Vempala Georgia Tech The Volume Problem Solution: Use Randomness! Reference Complexity Remarks Dyer, Frieze, and Kannan First polynomial algorithm (89) Lovsz and Simonovits (90) Isoperimetry,
SLIDE 2
SLIDE 3
Solution: Use Randomness!
- Reference
Complexity Remarks Dyer, Frieze, and Kannan (‘89) First polynomial algorithm Lovász and Simonovits (‘90) Isoperimetry, localization Applegate and Kannan (‘90) Logconcave sampling and integration Lovász (‘90) Ball walk Dyer and Frieze (‘91) Lovász and Simonovits (‘93) Kannan, Lovász, and Simonovits (‘97) Speedy walk, isotropy Lovász and Vempala (‘03) Annealing, hit-and-run Cousins and Vempala (‘15) Gaussian cooling
SLIDE 4
Compute Volume in Phases [LV03]
SLIDE 5
Cooling with Gaussians
- …
SLIDE 6
Computing the Ratios
SLIDE 7
Metropolis Ball Walk
SLIDE 8
KLS Conjecture
SLIDE 9
Mixing Time of Ball Walk
SLIDE 10
Gaussian Sampling and Isoperimetry
SLIDE 11
Gaussian Cooling Algorithm
- Even at this faster rate, variance of
ratio estimator remains a constant!!!
SLIDE 12
Sketch of Analysis
. . . . . . . . . . . .
SLIDE 13
SLIDE 14
Bounding the Variance
SLIDE 15
Localization Lemma
SLIDE 16
Further Theoretical Questions
- Illustration of one step of hit-and-run.
SLIDE 17
MATLAB Implementation
- Volume algorithm has been implemented in MATLAB, available for
download from MATLAB File Exchange
- “volume computation matlab”
- Can compute volume of 100 dimensional polytopes in ~10 minutes.