relaxations via statistical query complexity Based on V F Will Perkins Santosh Vempala On the Complexity of Random Satisfiability Problems with Planted Solutions STOC 2015 V F ID: 529046
Download Presentation The PPT/PDF document "Lower bounds against convex" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
Lower bounds against convex relaxationsvia statistical query complexity
Based on:V. F., Will Perkins, Santosh Vempala. On the Complexity of Random Satisfiability Problems with Planted Solutions. STOC 2015V. F., Cristobal Guzman, Santosh Vempala. Statistical Query Algorithms for Stochastic Convex Optimization. SODA 2017V. F. A General Characterization of the Statistical Query Complexity. arXiv 2016
Vitaly Feldman IBM Research – AlmadenSlide2
The planBoolean constraint satisfaction problems
Convex relaxationsComparison with lower bounds against LP/SDP hierarchies(Known) sign-rank lower bounds via SQ complexitySlide3
MAX-CSPs
-SATGiven:
, where clause is OR of
(possibly negated) variables Is satisfiable?
MAX--CSPFind
for
-
ary
predicates
-SAT refutation
If is satisfiable output YES.If is random output NO with prob : -clauses are chosen randomly and uniformly from of all -clauses Unsatisfiable w.h.p. for
Best poly-time algorithm
uses clauses [Goerdt,Krivelevich 01; Coja-Oglan,Goerdt,Lanka 07; Allen,O’Donnell,Witmer 15]Conjectured to be hard [Feige 02]
Slide4
Convex relaxation for MAX-CSPs
Objective-wise mapping:Denote
Which
allow such mappings?
Clause
over
Convex
function
over a convex body Refutation gap :
:
(
)
Slide5
Outline
Optimization of in the stochastic setting has low SQ complexity
Lower bound on statistical query complexity of stochastic -SAT refutation
Convexrelaxation
-
-
Convex optimization algorithmsYES: the support of is satisfiable Slide6
Lower bound example I
-Lipschitz convex optimization needs
-Lipschitz
convex optimization needs
For
and any convex
all convex funcs s.t. Then
For
and any convex
all convex funcs s.t.
Then
Slide7
Lower bound example II
General convex optimization needs
For
and any convex
all convex
funcs
over
with range
Then Slide8
Lower bounds from algorithms
-Lipschitz convex optimization needs
convex optimization
needs
General convex
optimization
needs
Random walks Center of gravityProjected gradient descentEntropic mirror descent(multiplicative weights) Slide9
Statistical queries [Kearns ‘93]
over
Slide10
Statistical queries [Kearns ‘93]
is tolerance of the query;
SQ algorithm
oracle
Problem
:
If exists
a SQ algorithm that
solves
using
queries to
Slide11
Statistical queries [Kearns ‘93]
is tolerance of the query;
SQ algorithm
oracle
Applications:
Noise-tolerant learning
[Kearns
93; …]
Private data analysis
[
Dinur,Nissim
03;
Blum,Dwork,McSherry,Nissim
05;
DMN,Smith
06
;
]
Distributed/low communication/memory ML
[
Ben-
David,Dichterman
98; Chu et al., 06;
Balcan,Blum,Fine,Mansour
12;
Steinhardt,G
.
Valiant,Wager
15;
F
. 16
]
Evolvability
[L. Valiant 06;
F
. 08; …]
Adaptive data analysis
[
Dwork
,
F
.,
Hardt,Pitassi,Reingold,Roth
14; …]
Slide12
Outline
Optimization of in the stochastic setting has low SQ complexity
Lower bound on SQ complexity of stochastic -SAT refutation
Convexrelaxation
-
-
Convex optimization algorithmsSlide13
Stochastic convex optimization (SCO)
Convex body Class of convex functions over
What is the SQ complexity of
?
:
Unknown
distribution
over
-minimize
over :Find s.t. Standard: Given i.i.d. samples SQ: Given
Slide14
SQ algorithms for
SCO
Reduction from an optimization oracle to SQ oracle
Direct analysis of an existing SCO algorithm
New/modified algorithm
Hard work
ReductionsSlide15
Zero-order/value oracle
𝜂-approximate value oracle for 𝑓 over Given
returns
,
If for all
then for any over , can simulate [P.Valiant ‘11]
Slide16
Corollaries
Known results for arbitrary and Ellipsoid-based:
queries to
[Nemirovski,Yudin 77;
Grotschel,Lovasz,Schrijver
88]
Random walks:
queries to
[Belloni,Liang,Narayanan,Rakhlin 15; F.,Perkins,Vempala 15] Corollary: For = {all convex funcs over with range } In high dimension weaker than full access/gradient oracle [Nemirovski,Yudin
‘77; Singer,Vondrak
‘15; Li,
Risteski ‘16]Slide17
First-order/gradient oracles
If then equivalent to
To implement
need to estimate
within
in
Assuming that
!
Global approximate gradient oracle of
over Given returns , s.t. for all
Slide18
Mean vector estimation
Easy case: Coordinate-wise estimation: for every , ask query
Let be the answer of
. Then
What about
?
Coordinate-wise estimation requires
In contrast,
samples suffice
Mean estimation in :Given distribution over Find s.t. , where
Slide19
Kashin’s representation [Lyubarskii
, Vershynin 10]Use coordinate-wise mean-estimation in Kashin’s representation: For every , ask query
to
.
Vectors
provide
Kashin’s
representation with level
if : tight frame: low dynamic range: and Corollary: Mean estimation in
can be solved using
queries to
Thm [LV 10]: There exists Kashin’s representation of level for and can be constructed efficiently
Slide20
Other norms
normsWhat about the general case?Always in
Mostly openDifferent from sample complexity for some hard to compute norms Slide21
Example corollaries
-Lipschitz SCO:
For any convex
all convex funcs s.t.
-Lipschitz SCO
:
For any
convex all convex funcs s.t.
Slide22
Outline
Optimization of in the stochastic setting has low SQ complexity
Lower bound on SQ complexity of stochastic -SAT refutationFrom SQ dimension to SQ complexityLower bound on SQ dimension of
-SAT
Convexrelaxation
-
-
Convex optimization algorithmsSlide23
Stochastic -SAT refutation
If s.t. the support of is satisfiable, output YESIf
, output NO with prob
Slide24
SQ dimension
One-vs-many decision problems:Let be a set distributions over and be a reference distribution
over
:
for an input distribution
decide if
Fixed-distribution PAC learning
[
Blum,Furst,Jackson,Kearns,Mansour,Rudich 95; …]General statistical problemsLower bounds [F.,Grigorescu,Reyzin,Vempala,Xiao 13; FPV 15]Characterization [F. 16] Slide25
If
then any algorithm that solves
given access to
requires
>
queries
SQ dimension
of
[
F
. 16]
Slide26
o
f-SAT refutation
is a degree-
(multilinear) polynomial of with constant term
Concentration properties of low-degree polynomials over
:
for all
,
Hard family of distributions:
uniform
over all
-clauses in which satisfies an odd number of literals
Thm:
-
-
Slide27
Outline
Optimization of in the stochastic setting has low SQ complexity
Lower bound on SQ complexity of stochastic -SAT refutation
Convexrelaxation
-
-
Convex optimization algorithmsSlide28
Comparison with known approaches
Same:Objective-wise relaxation to functions over a fixed Incomparable/complementary: Known SQ based
Linear functions
Convex functions:
is
a
polytope
with bounded number of facets
is any convex body.
is boundedAssumes mapping s.t. and gapAssumes an gap in optimization outcomes“Variance”/“Overfitting”“Bias”/“Model misspecification”“Variance”/“Overfitting”“Bias”/“Model misspecification”Sherali-Adams,SOS/Laserre hierarchies: [Grigoriev 01; Shoenebeck 08; Charikar,Makarychev,Makarychev 09; O’Donnell,Witmer 14]LP extended formulations:
[Chan,Lee,Raghavendra,Steurer 13;
Kothari,Meka,Raghavendra
16][Barak,Moitra ’16]Slide29
Sign-rank lower bounds via SQ complexity
Dimension complexity:Let be a set of -valued functions over is the lowest
such that exists a mapping such that:
exists
, such that
,
Define
Then
For a matrix
,
Corollary:
Proved
by Forster
[2001]
Halfspaces over
can
be PAC learned
in
[
Blum,Frieze,Kannan,Vempala
96]
Learning of
(parity functions)
not
in
[Kearns 93; BFJKMR 95]
Slide30
Conclusions
Convex relaxations fail for XOR constraint optimizationSQ complexity lower bounds bridge between algorithms and structural lower boundsExtensionsOther MAX--CSPsStronger -wise reductions [
F., Ghazi ‘17]Many open problems