Download presentation
Presentation is loading. Please wait.
Published byDominic Lambert Modified over 9 years ago
1
1 Distribution-free testing algorithms for monomials with a sublinear number of queries Elya Dolev & Dana Ron Tel-Aviv University
2
2 Property testing of (Boolean) functions (“standard/uniform” version) f : {0,1} n {0,1} - the tested function F - family of functions (e.g. linear functions) Given a dist. par. and query access to f f x f(x) If f F, then accept w.p. 2/3 If dist(f,F) > then reject w.p 2/3 where dist(f,F) = min g F {dist(f,g)} and dist(f,g) = Pr x U [f(x) g(x)]
3
3 Property testing of (Boolean) functions distribution-free version f : {0,1} n {0,1} - the tested function F - family of functions (e.g. linear functions) D - (unknown) underlying distribution Given a dist. par. , access to examples distributed by D and query access to f fx f(x) If f F, then accept w.p. 2/3 If dist D (f,F) > then reject w.p 2/3 where dist D (f,F) = min g F {dist D (f,g)} and dist D (f,g) = Pr x D [f(x) g(x)] xD Inspired by dist-free PAC learning model [Valiant]
4
4 (Dist-free) Testing and Learning Dist-free testing was initially considered in [Goldreich,Goldwasser,R]. Observed that testing is no harder than (proper) learning (in particular, dist-free+queries). Q1: When is standard/dist-free testing easier than learning? Q2: What is relation btwn complexity of standard and dist-free testing?
5
5 Testing and Learning Quite a few classes for which standard testing is easier than learning (under the unif. dist. + queries): Linear functions [Blum,Luby,Rubinfeld] Low-degree polynomials [Rubinfeld&Sudan] Singletons, monomials, small monotone DNF [Parnas,R,Samorodintsky] Monotone functions [Ergun,Kannan,Kumar,Rubinfeld,Viswanathan] [Dodis,Goldreich,Lehman,Raskhodnikova,R,Samorodintsky] Small juntas [Fischer,Kindler,R,Safra,Samorodintsky] Small decision lists, decision trees, DNF (general) [Diakonikolas,Lee,Matulef,Onak,Rubinfeld,Servedio,Wan] Linear thresh. functions [Matulef,O’Donnell,Rubinfeld,Servedio]... Fewer positive results for dist-free testing [Halevy,Kushilevtz]x2. Tends to be more challenging.
6
6 Background on distribution-free testing One of the main positive (and general) results: if class has standard tester and can be self-corrected, then have dist-free tester [Halevy&Kushilevtz]. In particular gives dist-free testers for linear functions and low-degree polynomials. What about other classes of interest (e.g., from learning point of view) which don’t have self- correctors?
7
7 Background on distribution-free testing What about other classes of interest? [Glasner&Servedio] considered question for monomials (monotone/general), decision lists, linear thresh. func. Prove that every dist-free tester must perform ((n/log(n)) 1/5 ) queries (for const. ), in contrast to standard testing of classes where there is no dependence on n (and poly on 1/ ). Shows that strong dependence on n is unavoidable, but can we get some sublinear dependence on n? (Dist-free learning + queries requires linear dependence [Turan])
8
8 Our Results We give a positive answer to the question for monomials – both monotone and general. The complexity of our dist-free testing algorithms is O(n 1/2 log(n)/ ).
9
9 Standard vs. dist-free testing of monomials When the underlying distribution is uniform (standard testing), if f is a k-monomial, then Pr[f(x)=1] = 2 -k, and so can effectively consider only monomials where k = O(log(1/ ))). This is not generally true in dist-free case. Specifically, lower bound of [GS] constructs functions that depend on many variables and underlying dist. D helps to “hide non-monomiality”. Note: dist-free testing for (monotone) k-monomials when k is fixed, can be done using exp(k) samples+queries (combine [PRS] and [HK])
10
10 Dist-free testing of monotone monomials Let MM denote the class of monotone monomials (over n variables). Def of the violation hypergraph H f of a function f: - Its vertex set is {0,1} n ; - Each (hyper)edge is a subset e={y 0,y 1,…,y t } where f(y 0 )=0 and f(y j )=1 for every j>0, such that there is no g in MM consistent with f on e. Example: y 0 =010, y 1 =011, y 2 =110 (f(y 0 )=0, f(y 1 )=f(y 2 )=1) x 1 or x 3 must be in monomial x 1 cannot be in monomial x 3 cannot be in monomial (Notation: Z(y)={i: y i =0} (
11
11 Dist-free testing of monotone monomials Def of the violation hypergraph H f of a function f : - Its vertex set is {0,1} n ; - Each (hyper)edge is a subset e={y 0,y 1,…,y t } where f(y 0 )=0 and f(y j )=1 for every j>0, s.t. By def, if f is in MM then no edges in H f. Lemma: If dist D (f,MM) > , then D(C) > for every vertex cover C of H f. Testing algorithm tries to find an edge in H f.
12
12 Dist-free testing of monotone monomials Testing algorithm tries to find an edge in H f. Notation: for Z [n], y(Z) has all coordinates in Z equal 0, and others 1 (e.g., y({1,3}) = 0101) Basic building block: procedure that given y f -1 (0) searches for index j s.t. y j =0 and f(y({j}))=0 (i.e. x j must be in monomial if f in MM). Procedure performs binary search. - Starts with Z = Z(y). - In each iteration partitions Z to two equal parts Z 1, Z 2, and queries y(Z 1 ) and y(Z 2 ). - Continues with Z i s.t.f(y(Z i ))=0 (if f(y(Z 1 ))=f(y(Z 2 ))=1 then {y(Z),y(Z 1 ),y(Z 2 )} is an edge so can reject) - Stops when |Z|=1. Z Z 1 Z 2
13
13 Dist-free testing of monotone monomials Testing algorithm for MM - Obtain sample T of (n 1/2 / ) points dist. D. - For each y in T s.t. f(y)=0 run search proc. on y. - If search failed for some y then reject (and halt). Otherwise, let J be union of all indices returned. - Obtain sample T’ of (n 1/2 / ) points dist. D. - If exists y’ in T’ s.t. f(y’)=1 and Z(y’) J then reject, o.w. accept. Found edge {y(Z),y(Z 1 ),y(Z 2 )} Found edge {y({j}),y’}
14
14 Dist-free testing of monotone monomials Testing algorithm for MM - Obtain sample T of (n 1/2 / ) points dist. D. - For each y in T s.t. f(y)=0 run search proc. on y. - If search failed for some y then reject (and halt). Otherwise, let J be union of all indices returned. - Obtain sample T’ of (n 1/2 / ) points dist. D. - If exists y’ in T’ s.t. f(y’)=1 and Z(y’) J then reject, o.w. accept. Query complexity of alg: |T|log(n)+|T’| = O(n 1/2 log(n)/ ) If f in MM, alg always accepts. If dist D (f,MM) > then prove that rejects w.p. 2/3. Prove contrapositive: If f is accepted w.p. > 1/3 then can construct vertex cover C of H f s.t. D(C) ≤ implying that dist D (f,MM) ≤ .
15
15 Dist-free testing of general monomials First, modify notion of violation hypergraph H f : each edge {y 0,y 1,…,y t } still satisfies f(y 0 )=0, f(y j )=1, j>0, but now, j>0 Z(y j ) Z(y 0 ) and j>0 O(y j ) O(y 0 ). Next, binary search is performed on y in f -1 (0) but “w.r.t.” w in f -1 (1). Search finds index j s.t. f(w’)=0 for w’ that differs from w only on j’th coordinate. (in monotone case, implicitly w = 1 n ). After performing search on O(n 1/2 / ) sample points in f -1 (0) (w.r.t. same w) and obtaining set J of “relevant indices”, take additional sample and see if contains y in f -1 (1) s.t. y j w j for some j in J.
16
16 Summary and Open problems Give sublinear (Õ(n 1/2 )) algorithms for dist-free testing of monotone/general monomials. (Alg for general monomials extends alg for monotone monomials.) Two natural questions: What is exact complexity of dist-free testing of monomials? (Lower bound of [GS] is (n 1/5 )) What about other classes studied by [GS]? (Decision lists and linear threshold functions.)
17
17 Thanks
18
18 Dist-free testing of monotone monomials If f is accepted w.p. > 1/3 then can construct vertex cover C of H f s.t. D(C) ≤ implying that dist D (f,MM) ≤ . First put in C all (very few) points y f -1 (0) for which binary search would fail. For each other y f -1 (0) let j(y) be index found by binary search (which is a det. proc.). For set J, let Y(J) = {y f -1 (1) & Z(y’) J }.
Similar presentations
© 2024 SlidePlayer.com. Inc.
All rights reserved.