Approximating Orthogonal Matrices with Effective Givens - - PowerPoint PPT Presentation

approximating orthogonal matrices with effective givens
SMART_READER_LITE
LIVE PREVIEW

Approximating Orthogonal Matrices with Effective Givens - - PowerPoint PPT Presentation

Approximating Orthogonal Matrices with Effective Givens Factorization Thomas Frerix Technical University of Munich joint work with Joan Bruna (NYU) Poster #164 Givens Factorization of Orthogonal Matrices 1 0 0 0


slide-1
SLIDE 1

Approximating Orthogonal Matrices with Effective Givens Factorization

Thomas Frerix

Technical University of Munich

joint work with Joan Bruna (NYU) Poster #164

slide-2
SLIDE 2

Givens Factorization of Orthogonal Matrices

G T(i, j, α) =         

1 ··· ··· ··· 0

. . . ... . . . . . . . . .

0 ··· cos(α) ··· − sin(α) ··· 0

. . . . . . ... . . . . . .

0 ··· sin(α) ··· cos(α) ··· 0

. . . . . . . . . ... . . .

0 ··· ··· ··· 1

        

slide-3
SLIDE 3

Givens Factorization of Orthogonal Matrices

G T(i, j, α) =         

1 ··· ··· ··· 0

. . . ... . . . . . . . . .

0 ··· cos(α) ··· − sin(α) ··· 0

. . . . . . ... . . . . . .

0 ··· sin(α) ··· cos(α) ··· 0

. . . . . . . . . ... . . .

0 ··· ··· ··· 1

         Exact Givens Factorization U = G1 . . . GN N = d(d − 1) 2

slide-4
SLIDE 4

Approximate Givens Factorization

Approximate Givens Factorization U ≈ G1 . . . GN N ≪ d(d − 1) 2 computationally hard problem

slide-5
SLIDE 5

Approximate Givens Factorization

Approximate Givens Factorization U ≈ G1 . . . GN N ≪ d(d − 1) 2 computationally hard problem Our Questions in this Context

  • 1. Which orthogonal matrices can be effectively approximated?

(not all of them)

slide-6
SLIDE 6

Approximate Givens Factorization

Approximate Givens Factorization U ≈ G1 . . . GN N ≪ d(d − 1) 2 computationally hard problem Our Questions in this Context

  • 1. Which orthogonal matrices can be effectively approximated?

(not all of them)

  • 2. Which principles are behind effective approximation algorithms?

(sparsity-inducing algorithms)

slide-7
SLIDE 7

Motivation: Unitary Basis Transform / FFT

Advantageous Setting Once computed, applied many times

slide-8
SLIDE 8

Motivation: Unitary Basis Transform / FFT

Advantageous Setting Once computed, applied many times Unitary Basis Transform FFT: O

  • d2

→ O

  • d log(d)
slide-9
SLIDE 9

Motivation: Unitary Basis Transform / FFT

Advantageous Setting Once computed, applied many times Unitary Basis Transform FFT: O

  • d2

→ O

  • d log(d)
  • Application: Graph Fourier Transform
slide-10
SLIDE 10

Which Matrices can be Effectively Approximated?

Theorem Let ǫ > 0. If N =o

  • d2/ log(d)
  • , then as d → ∞,

µ  

  • U ∈ U(d)
  • inf

G1...GN

U −

  • n

Gn2 ≤ ǫ   → 0 , where µ is the Haar measure over U(d).

slide-11
SLIDE 11

Which Matrices can be Effectively Approximated?

Theorem Let ǫ > 0. If N =o

  • d2/ log(d)
  • , then as d → ∞,

µ  

  • U ∈ U(d)
  • inf

G1...GN

U −

  • n

Gn2 ≤ ǫ   → 0 , where µ is the Haar measure over U(d).

  • proof is based on an ǫ-covering argument
  • suggests computational-to-statistical gap together with experimental results (details

at poster)

slide-12
SLIDE 12

K-planted Distribution over SO(d)

Sample U = G1 . . . GK

  • choose subspace (ik, jk) uniformly with replacement
  • choose rotation angle αk ∈ [0, 2π) uniformly
slide-13
SLIDE 13

K-planted Distribution over SO(d)

Sample U = G1 . . . GK

  • choose subspace (ik, jk) uniformly with replacement
  • choose rotation angle αk ∈ [0, 2π) uniformly

0.2 0.4 0.6 0.8 1 0.2 0.4 0.6 0.8 1

K/d log2(d) ||U||0/d2

256 512 1024

K-planted matrices quickly become dense

slide-14
SLIDE 14

Minimizing Sparsity-Inducing Norms over O(d)

G T

N . . . G T N U ≈ I

ˆ U = G1 . . . GN

slide-15
SLIDE 15

Minimizing Sparsity-Inducing Norms over O(d)

G T

N . . . G T N U ≈ I

ˆ U = G1 . . . GN Approximation criterion

  • U − ˆ

U

  • F,sym := min

P∈Pd

  • U − ˆ

UP

  • F
slide-16
SLIDE 16

Minimizing Sparsity-Inducing Norms over O(d)

G T

N . . . G T N U ≈ I

ˆ U = G1 . . . GN Approximation criterion

  • U − ˆ

U

  • F,sym := min

P∈Pd

  • U − ˆ

UP

  • F

Better functions to be minimized greedily? f (U) := d−1U1 = d−1

d

  • i,j=1
  • Uij
slide-17
SLIDE 17

Minimizing Sparsity-Inducing Norms over O(d)

G T

N . . . G T N U ≈ I

ˆ U = G1 . . . GN Approximation criterion

  • U − ˆ

U

  • F,sym := min

P∈Pd

  • U − ˆ

UP

  • F

Better functions to be minimized greedily? f (U) := d−1U1 = d−1

d

  • i,j=1
  • Uij
  • Non-convex greedy step
  • global optimum in O(d2) amortized time complexity
slide-18
SLIDE 18

Thank you

Poster #164 https://github.com/tfrerix/givens-factorization