On the Average-case Complexity of MCSP and Its Variants. Shuichi Hirahara (The University of Tokyo) Rahul Santhanam (University of Oxford)

Size: px
Start display at page:

Download "On the Average-case Complexity of MCSP and Its Variants. Shuichi Hirahara (The University of Tokyo) Rahul Santhanam (University of Oxford)"

Transcription

1 On the Average-case Complexity of MCSP and Its Variants Shuichi Hirahara (The University of Tokyo) Rahul Santhanam (University of Oxford) CCC Riga July 6, 2017

2 Minimum Circuit Size Problem (MCSP) Input Truth table T 0,1 2n of a function f: 0,1 n 0,1 Output Is there a circuit of size s that computes f? Size parameter s N Example: s = 5 Output: YES x 1 x 2 x 1 x f =

3 Brief History of MCSP Dates back to 1950s. [Trakhtenbrot s survey] Kabanets & Cai (2000) revived interest, based on natural proofs [Razborov & Rudich (1997)]. [ABKvMR06, AHMPS08, AD14, AHK15, HP15, MW15, HW16, CIKK16, IS17, AH17, IKV17] MCSP P under cryptographic assumptions. MCSP is not NP-hard under restricted reductions. Open Problem: Is MCSP NP-complete?

4 Average-case Complexity Parameterized MCSP[s] for s: N N Input Truth table T 0,1 2n of a function f: 0,1 n 0,1 Size parameter s N Output Is there a circuit of size s(n) that computes f? Consider the uniform distribution on 0,1 2n. Consider zero-error average-case complexity. (i.e. Algorithms output 0, 1, or? ) #YES instances = s O(s) 2 2n

5 Natural Proofs Useful Against SIZE s and Average-case Algorithms for MCSP[s] Natural proof: MCSP s An algorithm that 1. accepts most truth tables, and 2. rejects every truth table of an easy function. YES NO instances

6 Natural Proofs Useful Against SIZE s and Average-case Algorithms for MCSP[s] Natural proof: Claim: MCSP s An algorithm that 1. accepts most truth tables, and 2. rejects every truth table of an easy function. YES Rejects Accepts Natural Proof NO instances

7 Natural Proofs Useful Against SIZE s and Average-case Algorithms for MCSP[s] Natural proof: Claim: MCSP s An algorithm that 1. accepts most truth tables, and 2. rejects every truth table of an easy function. YES? 0 Zero-error Algorithm for MCSP s NO instances

8 Natural Proofs Useful Against SIZE s and Average-case Algorithms for MCSP[s] Natural proof: Claim: MCSP s An algorithm that 1. accepts most truth tables, and 2. rejects every truth table of an easy function. YES 1? 0 Zero-error Algorithm for MCSP s NO instances

9 Natural Proofs Useful Against SIZE s and Average-case Algorithms for MCSP[s] Natural proof: Claim: MCSP s An algorithm that 1. accepts most truth tables, and 2. rejects every truth table of an easy function. YES Natural Proof NO instances

10 Average-case Complexity of MCSP is More Intuitive In the setting of worst-case complexity, MCSP s 1 m p MCSP[s 2 ]? Not known In the setting of average-case complexity, MCSP s 1 MCSP[s 2 ] for s 1 s 2 This reduction is given by the identity map.

11 Outline 1. Pseudorandom self-reducibility of MCSP 2. Hardness of MKTP under Popular Average-Case Conjectures 3. Unconditional Lower Bounds for MCSP

12 Outline 1. Pseudorandom self-reducibility of MCSP 2. Hardness of MKTP under Popular Average-Case Conjectures 3. Unconditional Lower Bounds for MCSP

13 Random Self-reducibility L is (1-query) randomly self-reducible def Randomized poly-time machine Input: x 0,1 N Query q Oracle L Answer L(q) Output L x w.h.p. q is uniformly distributed on 0,1 N

14 Worst-case to Average-case Reduction L is randomly self-reducible, and algorithm solves L on average. algorithm solves L on every inputs. Input: x 0,1 N Query q U N The average-case algorithm Oracle L Answer L(q) Output L x w.h.p.

15 Worst-case Average-case for NP Theorem ([Feigenbaum & Fortnow 1993], [Bogdanov & Trevisan 2006]) NP-complete sets are not randomly self-reducible (unless PH collapses). If MCSP is randomly self-reducible, it provides strong evidence of non-np-hardness of MCSP.

16 Pseudorandom self-reducibility L is (1-query) pseudorandomly self-reducible def Randomized poly-time machine Input: x 0,1 N Query q Answer L(q) Oracle L Output L x w.h.p. q and U N are indistinguishable by SIZE(poly).

17 Worst-case to Average-case Reduction for Feasibly-on-Average Algorithms L is pseudorandomly self-reducible, and algorithm solves L on average and its error set can be decided in P. algorithm solves L on every inputs. Input: x 0,1 N Query q c U N E.g. A poly-time algorithm Oracle L Answer L(q) Output L x w.h.p.

18 MCSP is Pseudorandomly self-reducible Theorem Assume exponentially hard one-way functions exist. Then, for any s: N N, MCSP s n c, s + n c is pseudorandomly reducible to MCSP s.

19 MCSP is Pseudorandomly self-reducible Theorem Assume exponentially hard one-way functions exist. Then, for any s: N N, MCSP s n c, s + n c is pseudorandomly reducible to MCSP s. MCSP s n c, s + n c is the promise problem such that YES instances are truth tables of circuits of size s n n c, NO instances are truth tables of circuits of size > s n + n c.

20 MCSP is Pseudorandomly self-reducible Theorem Assume exponentially hard one-way functions exist. Then, for any s: N N, MCSP s n c, s + n c is pseudorandomly reducible to MCSP s. def f is exponentially hard one-way function. ε > 0 such that Pr x 0,1 n C f x f 1 f x < 2 nε for any circuit C of size < 2 nε.

21 MCSP is Pseudorandomly self-reducible Theorem Assume exponentially hard one-way functions exist. Then, for any s: N N, MCSP s n c, s + n c is pseudorandomly reducible to MCSP s. Main Ingredient: PseudoRandom Function Generator F (PRFG) F: 0,1 no(1) 0,1 2n is a PRFG def 1. F U n O 1 c U 2 n. (computationally indistinguishable) 2. The circuit complexity of F(r) is n c. A PRFG can be constructed from an exponentially hard OWF. [Razborov & Rudich 97], [Goldreich, Goldwasser & Micali 86]

22 Pseudorandom Self-reduction for MCSP Take a pseudorandom function generator F: 0,1 no 1 0,1 2n. Input: T 0,1 2n Pick r randomly. Query q T F(r) MCSP s oracle Output a Answer a 0,1 F r c U 2 n q = T F r c T U 2 n U 2 n. circuit complexity of T (circuit complexity of q) n c.

23 Pseudorandom self-reduction Summary of the 1 st part: 1. Introduced the notion of pseudorandom self-reduction. 2. MCSP is pseudorandomly self-reducible under a standard cryptographic assumption. Open Problem Are NP-complete sets pseudorandomly self-reducible under standard cryptographic assumptions?

24 Outline 1. Pseudorandom self-reducibility of MCSP 2. Hardness of MKTP under Popular Average-Case Conjectures 3. Unconditional Lower Bounds for MCSP

25 MKTP (Minimum Kolmogorov Time-bounded Complexity Problem) Input Output x 0,1 KT x s? Size parameter s N (Intuitively: Can each bit of x be described efficiently by a random access machine?) Definition of KT complexity [Allender, Buhrman, Koucký, van Melkebeek & Ronneburger 06] KT x min d + t U d i = x i in time t for all i. Fact [ABKvMR06]: KT x (circuit complexity of x)

26 Hardness Under Popular Conjectures Theorem 1. MKTP is Random 3SAT-hard (in the sense of Feige). 2. MKTP is Planted Clique-hard. 3. MKTP and MCSP are hard under Alekhnovich s hypothesis about linear equations with noise. Previously, SZK T BPP MCSP was known. [Allender & Das 2014] Our results give the first hardness results based on problems not known to be in SZK.

27 Random 3SAT [Feige 2002] Average-case version of 3SAT Distribution on inputs: A 3CNF formula with n variables and m = Δn clauses (Δ: a large constant) Each clause is chosen uniformly at random. Feige s Hypothesis (Random 3SAT is hard for P) There is no polynomial-time algorithm that 1. accepts every satisfiable formula, and 2. rejects most 3CNF formulae.

28 Random 3SAT hardness Theorem There is a poly-time algorithm with oracle access to MKTP that refutes Feige s hypothesis. Recently, Ryan O Donnell conjectured that Random 3SAT cannot be solved by even conp algorithms. In particular, his conjecture implies that MKTP is not in conp.

29 Proof of Random 3SAT Hardness Construct a many-one reduction: Random 3SAT φ MKTP (φ, θ) for some size parameter θ We need to claim: 1. KT φ > θ with high probability. (Most formulae are incompressible.) 2. KT φ θ for any satisfiable 3CNF formula φ. (Satisfiable formulae are quite rare instances.)

30 = Most Formulae are Incompressible Claim 1 KT φ > θ with high probability (over the choice of random 3CNF formula φ). log 8 n 3 random bits for each clause. (8 ways to negate variables.) φ = x 1 x 3 x 4 x 2 x 3 x 4 x 1 x 2 x 4 m log 8 n 3 random bits in total KT φ K φ > m log 8 n 3 O n w.h.p. θ

31 Satisfiable Formulae are Compressible Claim 2 KT φ m log 8 n 3 for any satisfiable 3CNF formula φ. Assume that some assignment a satisfies φ. Example: a = x 1 0, x 2 0, x 3 0, x 4 1. φ = x 1 x 3 x 4 x 2 x 3 x 4 x 1 x 2 x 4 This clause cannot appear in φ. Given a, each clause can be described by log 7 n bits of information. 3 (There are 7 ways to negate variables so that a clause is satisfied by a.) KT φ a + m log 7 n 3 m log 8 n 3 for large m.

32 Hardness Under Popular Conjectures Summary of the 2 nd part: 1. MKTP is Random 3SAT-hard, Planted Clique-hard, and hard under Alekhnovich s conjecture. 2. MCSP is hard under Alekhnovich s conjecture. Open Problem Is MCSP Random 3SAT-hard or Planted Clique-hard?

33 Outline 1. Pseudorandom self-reducibility of MCSP 2. Hardness of MKTP under Popular Average-Case Conjectures 3. Unconditional Lower Bounds for MCSP

34 Unconditional Lower Bounds Theorem For some size parameters s and s, 1. MKTP s cannot be decided with Ω 1 success on average by AC 0 p for any prime p. 2. MCSP s requires De Morgan formulae of size N 2 o 1 for input length N = 2 n. Proof idea Use ideas of pseudorandom generators for 1. AC 0 p [Fefferman, Shaltiel, Umans and Viola 13] 2. De Morgan formulae [Impagliazzo, Meka and Zuckerman 12]

35 MKTP AC 0 p on average Pseudorandom Generator for AC 0 [p] (p: odd prime) [FSUV 13] G: 0,1 n k 0,1 nk+k x 1,, x k x 1,, x k, PARITY x 1,, PARITY x k Fact. 1. KT G U nk nk + O n. 2. KT U nk+k nk + k n with high probability. MKTP oracle can be used to break G for k = n 2.

36 De Morgan Formula Lower Bounds Idea: Pseudorandom restriction A pseudorandom restriction shrinks a formula. Lemma. [Impagliazzo, Meka & Zuckerman 2012] There is a pseudorandom restriction ρ: N 0, 1, such that 1. ρ i = with probability p = N (1 o 1 ). 2. E ρ L f ρ p 2 L(f) for any function f. 3. Each bit of ρ can be computed in time N o 1. L MCSP s ρ pn for pseudorandom restriction ρ. (because MCSP s ρ depends on most unrestricted variables) pn L MCSP s ρ < p 2 L MCSP s.

37 Unconditional Lower Bounds Summary of the 3 rd part: 1. MKTP is not solvable by AC 0 [p] on average. 2. MCSP requires De Morgan Formulae of size N 2 o 1. Open Problems 1. MCSP AC 0 [p]? 2. MCSP requires De Morgan Formulae of size N 3 o 1?

38 Summary 1. MCSP is pseudorandomly self-reducible under a standard cryptographic assumption. 2. MKTP and MCSP are hard under popular averagecase complexity conjectures. 3. MKTP is not in AC 0 [p] on average. 4. MCSP requires De Morgan formulae of size N 2 o 1.

Limits of Minimum Circuit Size Problem as Oracle

Limits of Minimum Circuit Size Problem as Oracle Limits of Minimum Circuit Size Problem as Oracle Shuichi Hirahara(The University of Tokyo) Osamu Watanabe(Tokyo Institute of Technology) CCC 2016/05/30 Minimum Circuit Size Problem (MCSP) Input Truth table

More information

The Complexity of Complexity

The Complexity of Complexity The Complexity of Complexity New Insights on the (Non)hardness of Circuit Minimization and Related Problems Eric Allender Rutgers University Joint work with Shuichi Hirara (U. Tokyo) RodFest, January 5,

More information

Zero Knowledge and Circuit Minimization

Zero Knowledge and Circuit Minimization Zero Knowledge and Circuit Minimization Eric Allender 1 and Bireswar Das 2 1 Department of Computer Science, Rutgers University, USA allender@cs.rutgers.edu 2 IIT Gandhinagar, India bireswar@iitgn.ac.in

More information

Zero Knowledge and Circuit Minimization

Zero Knowledge and Circuit Minimization Electronic Colloquium on Computational Complexity, Report No. 68 (2014) Zero Knowledge and Circuit Minimization Eric Allender 1 and Bireswar Das 2 1 Department of Computer Science, Rutgers University,

More information

Some Results on Circuit Lower Bounds and Derandomization of Arthur-Merlin Problems

Some Results on Circuit Lower Bounds and Derandomization of Arthur-Merlin Problems Some Results on Circuit Lower Bounds and Derandomization of Arthur-Merlin Problems D. M. Stull December 14, 2016 Abstract We prove a downward separation for Σ 2 -time classes. Specifically, we prove that

More information

Some Results on Average-Case Hardness within the Polynomial Hierarchy

Some Results on Average-Case Hardness within the Polynomial Hierarchy Some Results on Average-Case Hardness within the Polynomial Hierarchy A. Pavan 1, Rahul Santhanam 2, and N. V. Vinodchandran 3 1 Department of Computer Science, Iowa State University 2 Department of Computer

More information

Conspiracies between Learning Algorithms, Lower Bounds, and Pseudorandomness

Conspiracies between Learning Algorithms, Lower Bounds, and Pseudorandomness Conspiracies between Learning Algorithms, Lower Bounds, and Pseudorandomness Igor Carboni Oliveira University of Oxford Joint work with Rahul Santhanam (Oxford) Context Minor algorithmic improvements imply

More information

Limitations of Efficient Reducibility to the Kolmogorov Random Strings

Limitations of Efficient Reducibility to the Kolmogorov Random Strings Limitations of Efficient Reducibility to the Kolmogorov Random Strings John M. HITCHCOCK 1 Department of Computer Science, University of Wyoming Abstract. We show the following results for polynomial-time

More information

Derandomizing from Random Strings

Derandomizing from Random Strings Derandomizing from Random Strings Harry Buhrman CWI and University of Amsterdam buhrman@cwi.nl Lance Fortnow Northwestern University fortnow@northwestern.edu Michal Koucký Institute of Mathematics, AS

More information

arxiv: v1 [cs.cc] 3 Nov 2016

arxiv: v1 [cs.cc] 3 Nov 2016 Conspiracies between Learning Algorithms, Circuit Lower Bounds and Pseudorandomness Igor C. Oliveira Charles University in Prague Rahul Santhanam University of Oxford arxiv:1611.01190v1 [cs.cc] 3 Nov 2016

More information

Identifying an Honest EXP NP Oracle Among Many

Identifying an Honest EXP NP Oracle Among Many Identifying an Honest EXP NP Oracle Among Many Shuichi Hirahara The University of Tokyo CCC 18/6/2015 Overview Dishonest oracle Honest oracle Queries Which is honest? Our Contributions Selector 1. We formulate

More information

CS 151 Complexity Theory Spring Solution Set 5

CS 151 Complexity Theory Spring Solution Set 5 CS 151 Complexity Theory Spring 2017 Solution Set 5 Posted: May 17 Chris Umans 1. We are given a Boolean circuit C on n variables x 1, x 2,..., x n with m, and gates. Our 3-CNF formula will have m auxiliary

More information

Parallel Repetition of Zero-Knowledge Proofs and the Possibility of Basing Cryptography on NP-Hardness

Parallel Repetition of Zero-Knowledge Proofs and the Possibility of Basing Cryptography on NP-Hardness Parallel Repetition of Zero-Knowledge Proofs and the Possibility of Basing Cryptography on NP-Hardness Rafael Pass Cornell University rafael@cs.cornell.edu January 29, 2007 Abstract Two long-standing open

More information

: On the P vs. BPP problem. 18/12/16 Lecture 10

: On the P vs. BPP problem. 18/12/16 Lecture 10 03684155: On the P vs. BPP problem. 18/12/16 Lecture 10 Natural proofs Amnon Ta-Shma and Dean Doron 1 Natural proofs The ultimate goal we have is separating classes (or proving they are equal if they are).

More information

Worst-Case to Average-Case Reductions Revisited

Worst-Case to Average-Case Reductions Revisited Worst-Case to Average-Case Reductions Revisited Dan Gutfreund 1 and Amnon Ta-Shma 2 1 SEAS, Harvard University, Cambridge, MA 02138 danny@eecs.harvard.edu 2 Computer Science Department, Tel-Aviv University,

More information

Majority is incompressible by AC 0 [p] circuits

Majority is incompressible by AC 0 [p] circuits Majority is incompressible by AC 0 [p] circuits Igor Carboni Oliveira Columbia University Joint work with Rahul Santhanam (Univ. Edinburgh) 1 Part 1 Background, Examples, and Motivation 2 Basic Definitions

More information

Questions Pool. Amnon Ta-Shma and Dean Doron. January 2, Make sure you know how to solve. Do not submit.

Questions Pool. Amnon Ta-Shma and Dean Doron. January 2, Make sure you know how to solve. Do not submit. Questions Pool Amnon Ta-Shma and Dean Doron January 2, 2017 General guidelines The questions fall into several categories: (Know). (Mandatory). (Bonus). Make sure you know how to solve. Do not submit.

More information

Length-Increasing Reductions for PSPACE-Completeness

Length-Increasing Reductions for PSPACE-Completeness Length-Increasing Reductions for PSPACE-Completeness John M. Hitchcock 1 and A. Pavan 2 1 Department of Computer Science, University of Wyoming. jhitchco@cs.uwyo.edu 2 Department of Computer Science, Iowa

More information

Meta-Algorithms vs. Circuit Lower Bounds Valentine Kabanets

Meta-Algorithms vs. Circuit Lower Bounds Valentine Kabanets Meta-Algorithms vs. Circuit Lower Bounds Valentine Kabanets Tokyo Institute of Technology & Simon Fraser University Understanding Efficiency Better understanding of Efficient Computation Good Algorithms

More information

Exponential time vs probabilistic polynomial time

Exponential time vs probabilistic polynomial time Exponential time vs probabilistic polynomial time Sylvain Perifel (LIAFA, Paris) Dagstuhl January 10, 2012 Introduction Probabilistic algorithms: can toss a coin polynomial time (worst case) probability

More information

Average-Case Complexity

Average-Case Complexity Foundations and Trends R in Theoretical Computer Science Vol. 2, No 1 (2006) 1 106 c 2006 A. Bogdanov and L. Trevisan DOI: 10.1561/0400000004 Average-Case Complexity Andrej Bogdanov 1 and Luca Trevisan

More information

Average-Case Complexity

Average-Case Complexity arxiv:cs/0606037v2 [cs.cc] 29 Sep 2006 Average-Case Complexity Andrej Bogdanov 1 Luca Trevisan 2 February 1, 2008 1 adib@ias.edu. DIMACS. Work partly done at the Institute for Advanced Study, Scool of

More information

On Worst-Case to Average-Case Reductions for NP Problems

On Worst-Case to Average-Case Reductions for NP Problems On Worst-Case to Average-Case Reductions for NP Problems Andrej Bogdanov Luca Trevisan January 27, 2005 Abstract We show that if an NP-complete problem has a non-adaptive self-corrector with respect to

More information

IF NP LANGUAGES ARE HARD ON THE WORST-CASE, THEN IT IS EASY TO FIND THEIR HARD INSTANCES

IF NP LANGUAGES ARE HARD ON THE WORST-CASE, THEN IT IS EASY TO FIND THEIR HARD INSTANCES IF NP LANGUAGES ARE HARD ON THE WORST-CASE, THEN IT IS EASY TO FIND THEIR HARD INSTANCES Dan Gutfreund, Ronen Shaltiel, and Amnon Ta-Shma Abstract. We prove that if NP BPP, i.e., if SAT is worst-case hard,

More information

Lecture 20: conp and Friends, Oracles in Complexity Theory

Lecture 20: conp and Friends, Oracles in Complexity Theory 6.045 Lecture 20: conp and Friends, Oracles in Complexity Theory 1 Definition: conp = { L L NP } What does a conp computation look like? In NP algorithms, we can use a guess instruction in pseudocode:

More information

CSC 2429 Approaches to the P versus NP Question. Lecture #12: April 2, 2014

CSC 2429 Approaches to the P versus NP Question. Lecture #12: April 2, 2014 CSC 2429 Approaches to the P versus NP Question Lecture #12: April 2, 2014 Lecturer: David Liu (Guest) Scribe Notes by: David Liu 1 Introduction The primary goal of complexity theory is to study the power

More information

The Complexity of Explicit Constructions

The Complexity of Explicit Constructions The Complexity of Explicit Constructions Rahul Santhanam University of Edinburgh rsanthan@inf.ed.ac.uk Abstract The existence of extremal combinatorial objects, such as Ramsey graphs and expanders, is

More information

Uniform Derandomization

Uniform Derandomization Uniform Derandomization Simulation of BPP, RP and AM under Uniform Assumptions A. Antonopoulos (N.T.U.A.) Computation and Reasoning Laboratory 1 Uniform Derandomization of BPP Main Theorem Proof: Step

More information

Hardness amplification proofs require majority

Hardness amplification proofs require majority Hardness amplification proofs require majority Emanuele Viola Columbia University Work done at Harvard, IAS, and Columbia Joint work with Ronen Shaltiel University of Haifa January 2008 Circuit lower bounds

More information

CSC 5170: Theory of Computational Complexity Lecture 9 The Chinese University of Hong Kong 15 March 2010

CSC 5170: Theory of Computational Complexity Lecture 9 The Chinese University of Hong Kong 15 March 2010 CSC 5170: Theory of Computational Complexity Lecture 9 The Chinese University of Hong Kong 15 March 2010 We now embark on a study of computational classes that are more general than NP. As these classes

More information

Umans Complexity Theory Lectures

Umans Complexity Theory Lectures Complexity Theory Umans Complexity Theory Lectures Lecture 1a: Problems and Languages Classify problems according to the computational resources required running time storage space parallelism randomness

More information

1 Randomized Computation

1 Randomized Computation CS 6743 Lecture 17 1 Fall 2007 1 Randomized Computation Why is randomness useful? Imagine you have a stack of bank notes, with very few counterfeit ones. You want to choose a genuine bank note to pay at

More information

CS21 Decidability and Tractability

CS21 Decidability and Tractability CS21 Decidability and Tractability Lecture 20 February 23, 2018 February 23, 2018 CS21 Lecture 20 1 Outline the complexity class NP NP-complete probelems: Subset Sum NP-complete problems: NAE-3-SAT, max

More information

On the NP-Completeness of the Minimum Circuit Size Problem

On the NP-Completeness of the Minimum Circuit Size Problem On the NP-Completeness of the Minimum Circuit Size Problem John M. Hitchcock Department of Computer Science University of Wyoming A. Pavan Department of Computer Science Iowa State University Abstract

More information

On Worst-Case to Average-Case Reductions for NP Problems

On Worst-Case to Average-Case Reductions for NP Problems On Worst-Case to Average-Case Reductions for NP Problems Andrej Bogdanov Luca Trevisan January 24, 2006 Abstract We show that if an NP-complete problem has a non-adaptive self-corrector with respect to

More information

Lecture 10: Learning DNF, AC 0, Juntas. 1 Learning DNF in Almost Polynomial Time

Lecture 10: Learning DNF, AC 0, Juntas. 1 Learning DNF in Almost Polynomial Time Analysis of Boolean Functions (CMU 8-859S, Spring 2007) Lecture 0: Learning DNF, AC 0, Juntas Feb 5, 2007 Lecturer: Ryan O Donnell Scribe: Elaine Shi Learning DNF in Almost Polynomial Time From previous

More information

: On the P vs. BPP problem. 30/12/2016 Lecture 11

: On the P vs. BPP problem. 30/12/2016 Lecture 11 03684155: On the P vs. BPP problem. 30/12/2016 Lecture 11 Promise problems Amnon Ta-Shma and Dean Doron 1 Definitions and examples In a promise problem, we are interested in solving a problem only on a

More information

If NP languages are hard on the worst-case then it is easy to find their hard instances

If NP languages are hard on the worst-case then it is easy to find their hard instances If NP languages are hard on the worst-case then it is easy to find their hard instances Dan Gutfreund School of Computer Science and Engineering, The Hebrew University of Jerusalem, Israel, 91904 danig@cs.huji.ac.il

More information

CSC 2429 Approaches to the P vs. NP Question and Related Complexity Questions Lecture 2: Switching Lemma, AC 0 Circuit Lower Bounds

CSC 2429 Approaches to the P vs. NP Question and Related Complexity Questions Lecture 2: Switching Lemma, AC 0 Circuit Lower Bounds CSC 2429 Approaches to the P vs. NP Question and Related Complexity Questions Lecture 2: Switching Lemma, AC 0 Circuit Lower Bounds Lecturer: Toniann Pitassi Scribe: Robert Robere Winter 2014 1 Switching

More information

Pseudorandom Generators and Typically-Correct Derandomization

Pseudorandom Generators and Typically-Correct Derandomization Pseudorandom Generators and Typically-Correct Derandomization Jeff Kinne 1, Dieter van Melkebeek 1, and Ronen Shaltiel 2 1 Department of Computer Sciences, University of Wisconsin-Madison, USA {jkinne,dieter}@cs.wisc.edu

More information

conp, Oracles, Space Complexity

conp, Oracles, Space Complexity conp, Oracles, Space Complexity 1 What s next? A few possibilities CS161 Design and Analysis of Algorithms CS254 Complexity Theory (next year) CS354 Topics in Circuit Complexity For your favorite course

More information

Umans Complexity Theory Lectures

Umans Complexity Theory Lectures Umans Complexity Theory Lectures Lecture 12: The Polynomial-Time Hierarchy Oracle Turing Machines Oracle Turing Machine (OTM): Deterministic multitape TM M with special query tape special states q?, q

More information

Algebrization: A New Barrier in Complexity Theory

Algebrization: A New Barrier in Complexity Theory Algebrization: A New Barrier in Complexity Theory Scott Aaronson MIT Avi Wigderson Institute for Advanced Study Abstract Any proof of P NP will have to overcome two barriers: relativization and natural

More information

A.Antonopoulos 18/1/2010

A.Antonopoulos 18/1/2010 Class DP Basic orems 18/1/2010 Class DP Basic orems 1 Class DP 2 Basic orems Class DP Basic orems TSP Versions 1 TSP (D) 2 EXACT TSP 3 TSP COST 4 TSP (1) P (2) P (3) P (4) DP Class Class DP Basic orems

More information

Definition: conp = { L L NP } What does a conp computation look like?

Definition: conp = { L L NP } What does a conp computation look like? Space Complexity 28 Definition: conp = { L L NP } What does a conp computation look like? In NP algorithms, we can use a guess instruction in pseudocode: Guess string y of x k length and the machine accepts

More information

CS294: Pseudorandomness and Combinatorial Constructions September 13, Notes for Lecture 5

CS294: Pseudorandomness and Combinatorial Constructions September 13, Notes for Lecture 5 UC Berkeley Handout N5 CS94: Pseudorandomness and Combinatorial Constructions September 3, 005 Professor Luca Trevisan Scribe: Gatis Midrijanis Notes for Lecture 5 In the few lectures we are going to look

More information

In Search of an Easy Witness: Exponential Time vs. Probabilistic Polynomial Time

In Search of an Easy Witness: Exponential Time vs. Probabilistic Polynomial Time In Search of an Easy Witness: Exponential Time vs. Probabilistic Polynomial Time Russell Impagliazzo Department of Computer Science University of California, San Diego La Jolla, CA 92093-0114 russell@cs.ucsd.edu

More information

Lecture 23: Alternation vs. Counting

Lecture 23: Alternation vs. Counting CS 710: Complexity Theory 4/13/010 Lecture 3: Alternation vs. Counting Instructor: Dieter van Melkebeek Scribe: Jeff Kinne & Mushfeq Khan We introduced counting complexity classes in the previous lecture

More information

An Introduction to Proof Complexity, Part II.

An Introduction to Proof Complexity, Part II. An Introduction to Proof Complexity, Part II. Pavel Pudlák Mathematical Institute, Academy of Sciences, Prague and Charles University, Prague Computability in Europe 2009, Heidelberg 2 Overview Part I.

More information

Arthur and Merlin as Oracles

Arthur and Merlin as Oracles Update Meeting on Algorithms and Complexity, IMSc 110308 Arthur and Merlin as Oracles Lecturer: Venkat Chakravarthy Scribe: Bireswar Das In this talk we will discuss about two results 1) BPP NP P pram

More information

Pseudo-Deterministic Proofs

Pseudo-Deterministic Proofs Pseudo-Deterministic Proofs Shafi Goldwasser 1, Ofer Grossman 2, and Dhiraj Holden 3 1 MIT, Cambridge MA, USA shafi@theory.csail.mit.edu 2 MIT, Cambridge MA, USA ofer.grossman@gmail.com 3 MIT, Cambridge

More information

Lecture 7 Limits on inapproximability

Lecture 7 Limits on inapproximability Tel Aviv University, Fall 004 Lattices in Computer Science Lecture 7 Limits on inapproximability Lecturer: Oded Regev Scribe: Michael Khanevsky Let us recall the promise problem GapCVP γ. DEFINITION 1

More information

CS151 Complexity Theory. Lecture 9 May 1, 2017

CS151 Complexity Theory. Lecture 9 May 1, 2017 CS151 Complexity Theory Lecture 9 Hardness vs. randomness We have shown: If one-way permutations exist then BPP δ>0 TIME(2 nδ ) ( EXP simulation is better than brute force, but just barely stronger assumptions

More information

Instance Compression for the Polynomial Hierarchy and Beyond

Instance Compression for the Polynomial Hierarchy and Beyond Instance Compression for the Polynomial Hierarchy and Beyond Chiranjit Chakraborty Rahul Santhanam School of Informatics, University of Edinburgh, UK Abstract. We define instance compressibility ([1],

More information

Notes for Lecture 2. Statement of the PCP Theorem and Constraint Satisfaction

Notes for Lecture 2. Statement of the PCP Theorem and Constraint Satisfaction U.C. Berkeley Handout N2 CS294: PCP and Hardness of Approximation January 23, 2006 Professor Luca Trevisan Scribe: Luca Trevisan Notes for Lecture 2 These notes are based on my survey paper [5]. L.T. Statement

More information

Foundations of Machine Learning and Data Science. Lecturer: Avrim Blum Lecture 9: October 7, 2015

Foundations of Machine Learning and Data Science. Lecturer: Avrim Blum Lecture 9: October 7, 2015 10-806 Foundations of Machine Learning and Data Science Lecturer: Avrim Blum Lecture 9: October 7, 2015 1 Computational Hardness of Learning Today we will talk about some computational hardness results

More information

CSC 5170: Theory of Computational Complexity Lecture 5 The Chinese University of Hong Kong 8 February 2010

CSC 5170: Theory of Computational Complexity Lecture 5 The Chinese University of Hong Kong 8 February 2010 CSC 5170: Theory of Computational Complexity Lecture 5 The Chinese University of Hong Kong 8 February 2010 So far our notion of realistic computation has been completely deterministic: The Turing Machine

More information

Lecture 7: Polynomial time hierarchy

Lecture 7: Polynomial time hierarchy Computational Complexity Theory, Fall 2010 September 15 Lecture 7: Polynomial time hierarchy Lecturer: Kristoffer Arnsfelt Hansen Scribe: Mads Chr. Olesen Recall that adding the power of alternation gives

More information

Proving SAT does not have Small Circuits with an Application to the Two Queries Problem

Proving SAT does not have Small Circuits with an Application to the Two Queries Problem Proving SAT does not have Small Circuits with an Application to the Two Queries Problem Lance Fortnow A. Pavan Samik Sengupta Abstract We show that if SAT does not have small circuits, then there must

More information

Making Hard Problems Harder

Making Hard Problems Harder Electronic Colloquium on Computational Complexity, Report No. 3 (2006) Making Hard Problems Harder Joshua Buresh-Oppenheim Simon Fraser University jburesho@cs.sfu.ca Rahul Santhanam Simon Fraser University

More information

Low-Depth Witnesses are Easy to Find

Low-Depth Witnesses are Easy to Find Low-Depth Witnesses are Easy to Find Luis Antunes U. Porto Lance Fortnow U. Chicago Alexandre Pinto U. Porto André Souto U. Porto Abstract Antunes, Fortnow, van Melkebeek and Vinodchandran captured the

More information

Towards NEXP versus BPP?

Towards NEXP versus BPP? Towards NEXP versus BPP? Ryan Williams Stanford University Abstract. We outline two plausible approaches to improving the miserable state of affairs regarding lower bounds against probabilistic polynomial

More information

THE COMPLEXITY OF CONSTRUCTING PSEUDORANDOM GENERATORS FROM HARD FUNCTIONS

THE COMPLEXITY OF CONSTRUCTING PSEUDORANDOM GENERATORS FROM HARD FUNCTIONS comput. complex. 13 (2004), 147 188 1016-3328/04/030147 42 DOI 10.1007/s00037-004-0187-1 c Birkhäuser Verlag, Basel 2004 computational complexity THE COMPLEXITY OF CONSTRUCTING PSEUDORANDOM GENERATORS

More information

On Basing Lower-Bounds for Learning on Worst-Case Assumptions

On Basing Lower-Bounds for Learning on Worst-Case Assumptions On Basing Lower-Bounds for Learning on Worst-Case Assumptions Benny Applebaum Boaz Barak David Xiao Abstract We consider the question of whether P NP implies that there exists some concept class that is

More information

A Short History of Computational Complexity

A Short History of Computational Complexity A Short History of Computational Complexity Lance Fortnow, Steve Homer Georgia Kaouri NTUAthens Overview 1936: Turing machine early 60 s: birth of computational complexity early 70 s: NP-completeness,

More information

Notes on Complexity Theory Last updated: October, Lecture 6

Notes on Complexity Theory Last updated: October, Lecture 6 Notes on Complexity Theory Last updated: October, 2015 Lecture 6 Notes by Jonathan Katz, lightly edited by Dov Gordon 1 PSPACE and PSPACE-Completeness As in our previous study of N P, it is useful to identify

More information

Journal of Computer and System Sciences

Journal of Computer and System Sciences Journal of Computer and System Sciences 77 (2011) 91 106 Contents lists available at ScienceDirect Journal of Computer and System Sciences www.elsevier.com/locate/jcss Infeasibility of instance compression

More information

1 AC 0 and Håstad Switching Lemma

1 AC 0 and Håstad Switching Lemma princeton university cos 522: computational complexity Lecture 19: Circuit complexity Lecturer: Sanjeev Arora Scribe:Tony Wirth Complexity theory s Waterloo As we saw in an earlier lecture, if PH Σ P 2

More information

MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS

MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS ERIC ALLENDER, JOSHUA A. GROCHOW, DIETER VAN MELKEBEEK, CRISTOPHER MOORE, AND ANDREW MORGAN Abstract. We study the computational power of deciding

More information

Complete problems for classes in PH, The Polynomial-Time Hierarchy (PH) oracle is like a subroutine, or function in

Complete problems for classes in PH, The Polynomial-Time Hierarchy (PH) oracle is like a subroutine, or function in Oracle Turing Machines Nondeterministic OTM defined in the same way (transition relation, rather than function) oracle is like a subroutine, or function in your favorite PL but each call counts as single

More information

Essential facts about NP-completeness:

Essential facts about NP-completeness: CMPSCI611: NP Completeness Lecture 17 Essential facts about NP-completeness: Any NP-complete problem can be solved by a simple, but exponentially slow algorithm. We don t have polynomial-time solutions

More information

The Minimum Oracle Circuit Size Problem

The Minimum Oracle Circuit Size Problem The Minimum Oracle Circuit Size Problem The MIT Faculty has made this article openly available. Please share how this access benefits you. Your story matters. Citation As Published Publisher Allender,

More information

Some Depth Two (and Three) Threshold Circuit Lower Bounds. Ryan Williams Stanford Joint work with Daniel Kane (UCSD)

Some Depth Two (and Three) Threshold Circuit Lower Bounds. Ryan Williams Stanford Joint work with Daniel Kane (UCSD) Some Depth Two (and Three) Threshold Circuit Lower Bounds Ryan Williams Stanford Joint wor with Daniel Kane (UCSD) Introduction Def. f n : 0,1 n 0,1 is a linear threshold function (LTF) if there are w

More information

Algorithms as Lower Bounds

Algorithms as Lower Bounds Algorithms as Lower Bounds Lecture 3 Part 1: Solving QBF and NC1 Lower Bounds (Joint work with Rahul Santhanam, U. Edinburgh) Part 2: Time-Space Tradeoffs for SAT Ryan Williams Stanford University Quantified

More information

PSEUDORANDOMNESS AND AVERAGE-CASE COMPLEXITY VIA UNIFORM REDUCTIONS

PSEUDORANDOMNESS AND AVERAGE-CASE COMPLEXITY VIA UNIFORM REDUCTIONS PSEUDORANDOMNESS AND AVERAGE-CASE COMPLEXITY VIA UNIFORM REDUCTIONS Luca Trevisan and Salil Vadhan Abstract. Impagliazzo and Wigderson (36th FOCS, 1998) gave the first construction of pseudorandom generators

More information

CS154, Lecture 17: conp, Oracles again, Space Complexity

CS154, Lecture 17: conp, Oracles again, Space Complexity CS154, Lecture 17: conp, Oracles again, Space Complexity Definition: conp = { L L NP } What does a conp computation look like? In NP algorithms, we can use a guess instruction in pseudocode: Guess string

More information

FOURIER CONCENTRATION FROM SHRINKAGE

FOURIER CONCENTRATION FROM SHRINKAGE FOURIER CONCENTRATION FROM SHRINKAGE Russell Impagliazzo and Valentine Kabanets March 29, 2016 Abstract. For a class F of formulas (general de Morgan or read-once de Morgan), the shrinkage exponent Γ F

More information

Circuit Lower Bounds for Nondeterministic Quasi-Polytime: An Easy Witness Lemma for NP and NQP

Circuit Lower Bounds for Nondeterministic Quasi-Polytime: An Easy Witness Lemma for NP and NQP Electronic Colloquium on Computational Complexity, Report No. 188 (2017) Circuit Lower Bounds for Nondeterministic Quasi-Polytime: An Easy Witness Lemma for NP and NQP Cody D. Murray MIT R. Ryan Williams

More information

Finish K-Complexity, Start Time Complexity

Finish K-Complexity, Start Time Complexity 6.045 Finish K-Complexity, Start Time Complexity 1 Kolmogorov Complexity Definition: The shortest description of x, denoted as d(x), is the lexicographically shortest string such that M(w) halts

More information

1 Distributional problems

1 Distributional problems CSCI 5170: Computational Complexity Lecture 6 The Chinese University of Hong Kong, Spring 2016 23 February 2016 The theory of NP-completeness has been applied to explain why brute-force search is essentially

More information

A Note on the Karp-Lipton Collapse for the Exponential Hierarchy

A Note on the Karp-Lipton Collapse for the Exponential Hierarchy A Note on the Karp-Lipton Collapse for the Exponential Hierarchy Chris Bourke Department of Computer Science & Engineering University of Nebraska Lincoln, NE 68503, USA Email: cbourke@cse.unl.edu January

More information

MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS

MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS SIAM J. COMPUT. Vol. 47, No. 4, pp. 1339 1372 c 2018 Society for Industrial and Applied Mathematics MINIMUM CIRCUIT SIZE, GRAPH ISOMORPHISM, AND RELATED PROBLEMS ERIC ALLENDER, JOSHUA A. GROCHOW, DIETER

More information

CS 880: Advanced Complexity Theory Final Project: Hardness Amplification within NP. Abstract

CS 880: Advanced Complexity Theory Final Project: Hardness Amplification within NP. Abstract CS 880: Advanced Compleity Theory Final Project: Hardness Amplification within NP Priyananda Shenoy May 2008 1 Introduction Abstract We survey the current status of the problem of Hardness amplification

More information

On the optimal compression of sets in P, NP, P/poly, PSPACE/poly

On the optimal compression of sets in P, NP, P/poly, PSPACE/poly On the optimal compression of sets in P, NP, P/poly, PSPACE/poly Marius Zimand Towson University CCR 2012- Cambridge Marius Zimand (Towson U.) Compression P, NP, P/poly sets 2011 1 / 19 The language compression

More information

Some Results on Derandomization

Some Results on Derandomization Some Results on Derandomization Harry Buhrman CWI and University of Amsterdam Lance Fortnow University of Chicago A. Pavan Iowa State University Abstract We show several results about derandomization including

More information

Two Comments on Targeted Canonical Derandomizers

Two Comments on Targeted Canonical Derandomizers Two Comments on Targeted Canonical Derandomizers Oded Goldreich Department of Computer Science Weizmann Institute of Science Rehovot, Israel. oded.goldreich@weizmann.ac.il April 8, 2011 Abstract We revisit

More information

Hardness Amplification for Errorless Heuristics

Hardness Amplification for Errorless Heuristics Hardness Amplification for Errorless Heuristics Andrej Bogdanov Muli Safra August 10, 2007 Abstract An errorless heuristic is an algorithm that on all inputs returns either the correct answer or the special

More information

Kernel-Size Lower Bounds. The Evidence from Complexity Theory

Kernel-Size Lower Bounds. The Evidence from Complexity Theory : The Evidence from Complexity Theory IAS Worker 2013, Warsaw Part 2/3 Note These slides are a slightly revised version of a 3-part tutorial given at the 2013 Workshop on Kernelization ( Worker ) at the

More information

1 Cryptographic hash functions

1 Cryptographic hash functions CSCI 5440: Cryptography Lecture 6 The Chinese University of Hong Kong 23 February 2011 1 Cryptographic hash functions Last time we saw a construction of message authentication codes (MACs) for fixed-length

More information

Unconditional Lower Bounds Against Advice

Unconditional Lower Bounds Against Advice Unconditional Lower Bounds Against Advice Harry Buhrman Lance Fortnow Rahul Santhanam Abstract We show several unconditional lower bounds for exponential time classes against polynomial time classes with

More information

Symposium on Theoretical Aspects of Computer Science 2008 (Bordeaux), pp

Symposium on Theoretical Aspects of Computer Science 2008 (Bordeaux), pp Symposium on Theoretical Aspects of Computer Science 2008 (Bordeaux), pp. 157-168 www.stacs-conf.org FINDING IRREFUTABLE CERTIFICATES FOR S p 2 ARTHUR AND MERLIN VIA VENKATESAN T. CHAKARAVARTHY AND SAMBUDDHA

More information

Does Looking Inside a Circuit Help?

Does Looking Inside a Circuit Help? Electronic Colloquium on Computational Complexity, Report No. 109 (2017) Does Looking Inside a Circuit Help? Russell Impagliazzo Valentine Kabanets Antonina Kolokolova Pierre McKenzie Shadab Romani June

More information

Hard instances for satisfiability and quasi-one-way functions

Hard instances for satisfiability and quasi-one-way functions Hard instances for satisfiability and quasi-one-way functions Andrej Bogdanov 1 Kunal Talwar 2 Andrew Wan 3 1 The Chinese University of Hong Kong 2 Microsoft Research SVC 3 Columbia University andrejb@cse.cuhk.edu.hk

More information

NONDETERMINISTIC CIRCUIT LOWER BOUNDS FROM MILDLY DERANDOMIZING ARTHUR-MERLIN GAMES

NONDETERMINISTIC CIRCUIT LOWER BOUNDS FROM MILDLY DERANDOMIZING ARTHUR-MERLIN GAMES NONDETERMINISTIC CIRCUIT LOWER BOUNDS FROM MILDLY DERANDOMIZING ARTHUR-MERLIN GAMES Barış Aydınlıoğlu and Dieter van Melkebeek March 24, 2014 Abstract. In several settings derandomization is known to follow

More information

Lecture 9: Polynomial-Time Hierarchy, Time-Space Tradeoffs

Lecture 9: Polynomial-Time Hierarchy, Time-Space Tradeoffs CSE 531: Computational Complexity I Winter 2016 Lecture 9: Polynomial-Time Hierarchy, Time-Space Tradeoffs Feb 3, 2016 Lecturer: Paul Beame Scribe: Paul Beame 1 The Polynomial-Time Hierarchy Last time

More information

On Interactive Proofs with a Laconic Prover

On Interactive Proofs with a Laconic Prover On Interactive Proofs with a Laconic Prover Oded Goldreich Salil Vadhan Avi Wigderson February 11, 2003 Abstract We continue the investigation of interactive proofs with bounded communication, as initiated

More information

1 Cryptographic hash functions

1 Cryptographic hash functions CSCI 5440: Cryptography Lecture 6 The Chinese University of Hong Kong 24 October 2012 1 Cryptographic hash functions Last time we saw a construction of message authentication codes (MACs) for fixed-length

More information

Hardness Amplification Proofs Require Majority

Hardness Amplification Proofs Require Majority Hardness Amplification Proofs Require Majority Ronen Shaltiel Emanuele Viola April 28, 2010 Abstract Hardness amplification is the fundamental task of converting a δ-hard function f : {0, 1} n {0, 1} into

More information

A note on exponential circuit lower bounds from derandomizing Arthur-Merlin games

A note on exponential circuit lower bounds from derandomizing Arthur-Merlin games Electronic Colloquium on Computational Complexity, Report No. 74 (200) A note on exponential circuit lower bounds from derandomizing Arthur-Merlin games Harry Buhrman Scott Aaronson MIT aaronson@csail.mit.edu

More information

Article begins on next page

Article begins on next page Minimum Circuit Size, Graph Isomorphism, and Related Problems Rutgers University has made this article freely available. Please share how this access benefits you. Your story matters. [https://rucore.libraries.rutgers.edu/rutgers-lib/56543/story/]

More information