DOCSLIB.ORG

P, NP, and NP-Completeness: the Basics of Computational Complexity

Total Page：16

File Type：pdf, Size：1020Kb

Download full-text PDF Read full-text

Load more

P NP and NPCompleteness The Basics of Computational Complexity Oded Goldreich Department of Computer Science and Applied Mathematics Weizmann Institute of Science Rehovot Israel June I to Dana c Copyright by Oded Goldreich Permission to make copies of part or all of this work for p ersonal or classro om use is granted without fee provided that copies are not made or distributed for prot or com mercial advantage and that new copies b ear this notice and the full citation on the rst page Abstracting with credit is p ermitted II Preface The strive for eciency is ancient and universal as time and other resources are always in shortage Thus the question of which tasks can b e p erformed eciently is central to the human exp erience A key step towards the systematic study of the aforementioned question is a rigorous denition of the notion of a task and of pro cedures for solving tasks These denitions were provided by computability theory which emerged in the s This theory fo cuses on computational tasks and considers automated pro cedures ie computing devices and algorithms that may solve such tasks In fo cusing attention on computational tasks and algorithms computability theory has set the stage for the study of the computational resources like time that are required by such algorithms When this study fo cuses on the resources that are necessary for any algorithm that solves a particular task or a task of a particular type the study b ecomes part of the theory of Computational Complexity also known as Complexity Theory Complexity Theory is a central eld of the theoretical foundations of Computer Science It is concerned with the study of the intrinsic complexity of computational tasks That is a typical Complexity theoretic study refers to the computational resources required to solve a computational task or a class of such tasks rather than referring to a sp ecic algorithm or an algorithmic schema Actually research in Complexity Theory tends to start with and focus on the computational resources themselves and addresses the eect of limiting these resources on the class of tasks that can b e solved Thus Computational Complexity is the general study of the what can b e achieved within limited time andor other limited natural computational resources The most famous question of complexity theory is the PvsNP Question and the current b o ok is fo cused on it The PvsNP Question can b e phrased as asking whether or not nding solutions is harder than checking the correctness of solu tions An alternative formulation asks whether or not discovering pro ofs is harder than verifying their correctness that is is proving harder than verifying The fun In contrast when the fo cus is on the design and analysis of sp ecic algorithms rather than on the intrinsic complexity of the task the study b ecomes part of a related subeld that may b e called Algorithmic Design and Analysis Furthermore Algorithmic Design and Analysis tends to b e subdivided according to the domain of mathematics science and engineering in which the computational tasks arise In contrast Complexity Theory typically maintains a unity of the study of tasks solvable within certain resources regardless of the origins of these tasks III IV damental nature of this question is evident in each of these formulations which are in fact equivalent It is widely b elieved that the answer to these equivalent for mulations is that nding resp proving is harder than checking resp verifying that is it is b elieved that P is dierent from NP At present when faced with a seemingly hard problem in NP we can only hop e to prove that it is not in P assuming that NP is dierent from P This is where the theory of NPcompleteness which is based on the notion of an ecient reduction comes into the picture In general one computational problem is e ciently reducible to another problem if it is p ossible to eciently solve the former when provided with an ecient algorithm for solving the latter A problem in NP is NPcomplete if any problem in NP is eciently reducible to it Amazingly enough NPcomplete problems exist and furthermore hundreds of natural compu tational problems arising in many dierent areas of mathematics and science are NPcomplete The main fo cus of the current b o ok is on the PvsNP Question and the theory of NPcompleteness Additional topics that are covered include the treatment of the general notion of an ecient reduction b etween computational problems which provides a tighter relation b etween the aforementioned search and decision problems The b o ok also provides adequate preliminaries regarding computational problems and computational mo dels Relation to a dierent b o ok of the author The current b o ok is a revision of Chapter and Section of the authors b o ok Computational Complexity A Con ceptual Perspective The revision was aimed at making the b o ok more friendly to the novice In particular several pro ofs were further detailed and numerous exercises were added Website for notices regarding this b o ok We intend to maintain a website listing corrections of various types The lo cation of the site is httpwwwwisdomweizmannacilodedbcbookhtml Overview This b o ok starts by providing the relevant background on computability theory which is the setting in which complexity theoretic questions are b eing studied Most imp ortantly this preliminary chapter ie Chapter provides a treatment of central notions such as search and decision problems algorithms that solve such problems and their complexity Sp ecial attention is given to the notion of a universal algorithm The main part of this b o ok ie Chapters is fo cused on the PvsNP Question and on the theory of NPcompleteness Additional topics covered in this part include the general notion of an ecient reduction with a sp ecial emphasis on selfreducibility the existence of problems in NP that are neither NPcomplete nor in P the class coNP optimal search algorithms and promise problems A brief overview of this main part follows Lo osely sp eaking the PvsNP Question refers to search problems for which the correctness of solutions can b e eciently checked ie there is an ecient algorithm that given a solution to a given instance determines whether or not the solution is correct Such search problems corresp ond to the class NP and the question is whether or not all these search problems can b e solved eciently ie is there an ecient algorithm that given an instance nds a correct solution Thus the Pvs NP Question can b e phrased as asking whether or not nding solutions is harder than checking the correctness of solutions An alternative formulation in terms of decision problems refers to assertions that have eciently veriable pro ofs of relatively short length Such sets of assertions corresp ond to the class NP and the question is whether or not pro ofs for such assertions can b e found eciently ie is there an ecient algorithm that given an assertion determines its validity andor nds a pro of for its validity Thus the PvsNP Question can b e phrased as asking whether or not discovering proofs is harder than verifying their correctness that is is proving harder than verifying or are pro ofs valuable at all Indeed it is widely b elieved that the answer to the two equivalent formulations is that nding resp discovering is harder than checking resp verifying that is that P is dierent than NP The fact that this natural conjecture is unsettled seems to b e one of the big sources of frustration of complexity theory The authors opinion however is that this feeling of frustration is out of place In any case at present when faced with a seemingly hard problem in NP we cannot exp ect to V VI prove that the problem is not in P unconditionally The b est we can exp ect is a conditional pro of that the said problem is not in P based on the assumption that NP is dierent from P The contrapositive is proving that if the said problem is in P then so is any problem in NP ie NP equals P This is where the theory of NPcompleteness comes into the picture The theory of NPcompleteness is based on the notion of an ecient reduction which is a relation b etween computational problems Lo osely sp eaking one com putational problem is eciently reducible to another problem if it is p ossible to eciently solve the former when provided with an ecient algorithm for solving the latter Thus the rst problem is not harder to solve than the second one A problem in NP is NPcomplete if any problem in NP is eciently reducible to it Thus the fate of the entire class NP with resp ect to inclusion in P rests with each individual NPcomplete problem In particular showing that a problem is NPcomplete implies that this problem is not in P unless NP equals P Amazingly enough NPcomplete problems exist and furthermore hundreds of natural compu tational problems arising in many dierent areas of mathematics and science are NPcomplete The foregoing paragraphs refer to material that is covered in Chapters Sp ecically Chapter is devoted to the PvsNP Question p er se Chapter is devoted to the notion of an ecient reduction and Chapter is devoted to the theory of NPcompleteness We mention that that NPcomplete problems are not the only seemingly hard problems in NP that is if P is dierent than NP then NP contains problems that are neither NPcomplete nor in P see Section Additional related topics are discussed in Chapter In particular in Sec tion it is shown that the PvsNP Question is not ab out inventing sophisticated algorithms or ruling out their existence but rather b oils down to the analysis of
Recommended publications
  • Complexity Theory Lecture 9 Co-NP Co-NP-Complete

    Complexity Theory Lecture 9 Co-NP Co-NP-Complete

    Complexity Theory 1 Complexity Theory 2 co-NP Complexity Theory Lecture 9 As co-NP is the collection of complements of languages in NP, and P is closed under complementation, co-NP can also be characterised as the collection of languages of the form: ′ L = x y y <p( x ) R (x, y) { |∀ | | | | → } Anuj Dawar University of Cambridge Computer Laboratory NP – the collection of languages with succinct certificates of Easter Term 2010 membership. co-NP – the collection of languages with succinct certificates of http://www.cl.cam.ac.uk/teaching/0910/Complexity/ disqualification. Anuj Dawar May 14, 2010 Anuj Dawar May 14, 2010 Complexity Theory 3 Complexity Theory 4 NP co-NP co-NP-complete P VAL – the collection of Boolean expressions that are valid is co-NP-complete. Any language L that is the complement of an NP-complete language is co-NP-complete. Any of the situations is consistent with our present state of ¯ knowledge: Any reduction of a language L1 to L2 is also a reduction of L1–the complement of L1–to L¯2–the complement of L2. P = NP = co-NP • There is an easy reduction from the complement of SAT to VAL, P = NP co-NP = NP = co-NP • ∩ namely the map that takes an expression to its negation. P = NP co-NP = NP = co-NP • ∩ VAL P P = NP = co-NP ∈ ⇒ P = NP co-NP = NP = co-NP • ∩ VAL NP NP = co-NP ∈ ⇒ Anuj Dawar May 14, 2010 Anuj Dawar May 14, 2010 Complexity Theory 5 Complexity Theory 6 Prime Numbers Primality Consider the decision problem PRIME: Another way of putting this is that Composite is in NP.
  • 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 6⊆ BPP, i.e., if SAT is worst-case hard, then for every probabilistic polynomial-time algorithm trying to decide SAT, there exists some polynomially samplable distribution that is hard for it. That is, the algorithm often errs on inputs from this distribution. This is the ¯rst worst-case to average-case reduction for NP of any kind. We stress however, that this does not mean that there exists one ¯xed samplable distribution that is hard for all probabilistic polynomial-time algorithms, which is a pre-requisite assumption needed for one-way func- tions and cryptography (even if not a su±cient assumption). Neverthe- less, we do show that there is a ¯xed distribution on instances of NP- complete languages, that is samplable in quasi-polynomial time and is hard for all probabilistic polynomial-time algorithms (unless NP is easy in the worst case). Our results are based on the following lemma that may be of independent interest: Given the description of an e±cient (probabilistic) algorithm that fails to solve SAT in the worst case, we can e±ciently generate at most three Boolean formulae (of increasing lengths) such that the algorithm errs on at least one of them. Keywords. Average-case complexity, Worst-case to average-case re- ductions, Foundations of cryptography, Pseudo classes Subject classi¯cation. 68Q10 (Modes of computation (nondetermin- istic, parallel, interactive, probabilistic, etc.) 68Q15 Complexity classes (hierarchies, relations among complexity classes, etc.) 68Q17 Compu- tational di±culty of problems (lower bounds, completeness, di±culty of approximation, etc.) 94A60 Cryptography 2 Gutfreund, Shaltiel & Ta-Shma 1.
  • Randomised Computation 1 TM Taking Advices 2 Karp-Lipton Theorem

    Randomised Computation 1 TM Taking Advices 2 Karp-Lipton Theorem

    INFR11102: Computational Complexity 29/10/2019 Lecture 13: More on circuit models; Randomised Computation Lecturer: Heng Guo 1 TM taking advices An alternative way to characterize P=poly is via TMs that take advices. Definition 1. For functions F : N ! N and A : N ! N, the complexity class DTime[F ]=A consists of languages L such that there exist a TM with time bound F (n) and a sequence fangn2N of “advices” satisfying: • janj ≤ A(n); • for jxj = n, x 2 L if and only if M(x; an) = 1. The following theorem explains the notation P=poly, namely “polynomial-time with poly- nomial advice”. S c Theorem 1. P=poly = c;d2N DTime[n ]=nd . Proof. If L 2 P=poly, then it can be computed by a family C = fC1;C2; · · · g of Boolean circuits. Let an be the description of Cn, andS the polynomial time machine M just reads 2 c this description and simulates it. Hence L c;d2N DTime[n ]=nd . For the other direction, if a language L can be computed in polynomial-time with poly- nomial advice, say by TM M with advices fang, then we can construct circuits fDng to simulate M, as in the theorem P ⊂ P=poly in the last lecture. Hence, Dn(x; an) = 1 if and only if x 2 L. The final circuit Cn just does exactly what Dn does, except that Cn “hardwires” the advice an. Namely, Cn(x) := Dn(x; an). Hence, L 2 P=poly. 2 Karp-Lipton Theorem Dick Karp and Dick Lipton showed that NP is unlikely to be contained in P=poly [KL80].
  • Computational Complexity and Intractability: an Introduction to the Theory of NP Chapter 9 2 Objectives

    Computational Complexity and Intractability: an Introduction to the Theory of NP Chapter 9 2 Objectives

    1 Computational Complexity and Intractability: An Introduction to the Theory of NP Chapter 9 2 Objectives . Classify problems as tractable or intractable . Define decision problems . Define the class P . Define nondeterministic algorithms . Define the class NP . Define polynomial transformations . Define the class of NP-Complete 3 Input Size and Time Complexity . Time complexity of algorithms: . Polynomial time (efficient) vs. Exponential time (inefficient) f(n) n = 10 30 50 n 0.00001 sec 0.00003 sec 0.00005 sec n5 0.1 sec 24.3 sec 5.2 mins 2n 0.001 sec 17.9 mins 35.7 yrs 4 “Hard” and “Easy” Problems . “Easy” problems can be solved by polynomial time algorithms . Searching problem, sorting, Dijkstra’s algorithm, matrix multiplication, all pairs shortest path . “Hard” problems cannot be solved by polynomial time algorithms . 0/1 knapsack, traveling salesman . Sometimes the dividing line between “easy” and “hard” problems is a fine one. For example, . Find the shortest path in a graph from X to Y (easy) . Find the longest path (with no cycles) in a graph from X to Y (hard) 5 “Hard” and “Easy” Problems . Motivation: is it possible to efficiently solve “hard” problems? Efficiently solve means polynomial time solutions. Some problems have been proved that no efficient algorithms for them. For example, print all permutation of a number n. However, many problems we cannot prove there exists no efficient algorithms, and at the same time, we cannot find one either. 6 Traveling Salesperson Problem . No algorithm has ever been developed with a Worst-case time complexity better than exponential .
  • 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∗ A. Pavany Department of Computer Science Department of Computer Science University of Wyoming Iowa State University Abstract We study the Minimum Circuit Size Problem (MCSP): given the truth-table of a Boolean function f and a number k, does there exist a Boolean circuit of size at most k computing f? This is a fundamental NP problem that is not known to be NP-complete. Previous work has studied consequences of the NP-completeness of MCSP. We extend this work and consider whether MCSP may be complete for NP under more powerful reductions. We also show that NP-completeness of MCSP allows for amplification of circuit complexity. We show the following results. • If MCSP is NP-complete via many-one reductions, the following circuit complexity amplifi- Ω(1) cation result holds: If NP\co-NP requires 2n -size circuits, then ENP requires 2Ω(n)-size circuits. • If MCSP is NP-complete under truth-table reductions, then EXP 6= NP \ SIZE(2n ) for some > 0 and EXP 6= ZPP. This result extends to polylog Turing reductions. 1 Introduction Many natural NP problems are known to be NP-complete. Ladner's theorem [14] tells us that if P is different from NP, then there are NP-intermediate problems: problems that are in NP, not in P, but also not NP-complete. The examples arising out of Ladner's theorem come from diagonalization and are not natural. A canonical candidate example of a natural NP-intermediate problem is the Graph Isomorphism (GI) problem.
  • Computational Complexity: a Modern Approach

    Computational Complexity: a Modern Approach

    i Computational Complexity: A Modern Approach Draft of a book: Dated January 2007 Comments welcome! Sanjeev Arora and Boaz Barak Princeton University [email protected] Not to be reproduced or distributed without the authors’ permission This is an Internet draft. Some chapters are more finished than others. References and attributions are very preliminary and we apologize in advance for any omissions (but hope you will nevertheless point them out to us). Please send us bugs, typos, missing references or general comments to [email protected] — Thank You!! DRAFT ii DRAFT Chapter 9 Complexity of counting “It is an empirical fact that for many combinatorial problems the detection of the existence of a solution is easy, yet no computationally efficient method is known for counting their number.... for a variety of problems this phenomenon can be explained.” L. Valiant 1979 The class NP captures the difficulty of finding certificates. However, in many contexts, one is interested not just in a single certificate, but actually counting the number of certificates. This chapter studies #P, (pronounced “sharp p”), a complexity class that captures this notion. Counting problems arise in diverse fields, often in situations having to do with estimations of probability. Examples include statistical estimation, statistical physics, network design, and more. Counting problems are also studied in a field of mathematics called enumerative combinatorics, which tries to obtain closed-form mathematical expressions for counting problems. To give an example, in the 19th century Kirchoff showed how to count the number of spanning trees in a graph using a simple determinant computation. Results in this chapter will show that for many natural counting problems, such efficiently computable expressions are unlikely to exist.
  • NP-Completeness Part I

    NP-Completeness Part I

    NP-Completeness Part I Outline for Today ● Recap from Last Time ● Welcome back from break! Let's make sure we're all on the same page here. ● Polynomial-Time Reducibility ● Connecting problems together. ● NP-Completeness ● What are the hardest problems in NP? ● The Cook-Levin Theorem ● A concrete NP-complete problem. Recap from Last Time The Limits of Computability EQTM EQTM co-RE R RE LD LD ADD HALT ATM HALT ATM 0*1* The Limits of Efficient Computation P NP R P and NP Refresher ● The class P consists of all problems solvable in deterministic polynomial time. ● The class NP consists of all problems solvable in nondeterministic polynomial time. ● Equivalently, NP consists of all problems for which there is a deterministic, polynomial-time verifier for the problem. Reducibility Maximum Matching ● Given an undirected graph G, a matching in G is a set of edges such that no two edges share an endpoint. ● A maximum matching is a matching with the largest number of edges. AA maximummaximum matching.matching. Maximum Matching ● Jack Edmonds' paper “Paths, Trees, and Flowers” gives a polynomial-time algorithm for finding maximum matchings. ● (This is the same Edmonds as in “Cobham- Edmonds Thesis.) ● Using this fact, what other problems can we solve? Domino Tiling Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling Solving Domino Tiling The Setup ● To determine whether you can place at least k dominoes on a crossword grid, do the following: ● Convert the grid into a graph: each empty cell is a node, and any two adjacent empty cells have an edge between them.
  • The Complexity Zoo

    The Complexity Zoo

    The Complexity Zoo Scott Aaronson www.ScottAaronson.com LATEX Translation by Chris Bourke [email protected] 417 classes and counting 1 Contents 1 About This Document 3 2 Introductory Essay 4 2.1 Recommended Further Reading ......................... 4 2.2 Other Theory Compendia ............................ 5 2.3 Errors? ....................................... 5 3 Pronunciation Guide 6 4 Complexity Classes 10 5 Special Zoo Exhibit: Classes of Quantum States and Probability Distribu- tions 110 6 Acknowledgements 116 7 Bibliography 117 2 1 About This Document What is this? Well its a PDF version of the website www.ComplexityZoo.com typeset in LATEX using the complexity package. Well, what’s that? The original Complexity Zoo is a website created by Scott Aaronson which contains a (more or less) comprehensive list of Complexity Classes studied in the area of theoretical computer science known as Computa- tional Complexity. I took on the (mostly painless, thank god for regular expressions) task of translating the Zoo’s HTML code to LATEX for two reasons. First, as a regular Zoo patron, I thought, “what better way to honor such an endeavor than to spruce up the cages a bit and typeset them all in beautiful LATEX.” Second, I thought it would be a perfect project to develop complexity, a LATEX pack- age I’ve created that defines commands to typeset (almost) all of the complexity classes you’ll find here (along with some handy options that allow you to conveniently change the fonts with a single option parameters). To get the package, visit my own home page at http://www.cse.unl.edu/~cbourke/.
  • University Microfilms International 300 North Zeeb Road Ann Arbor, Michigan 48106 USA St John's Road

    University Microfilms International 300 North Zeeb Road Ann Arbor, Michigan 48106 USA St John's Road

    INFORMATION TO USERS This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependant upon the quality of the original submitted. The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction. 1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity. 2. When an image on the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You w ill find a good image of the page in the adjacent frame. 3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete. 4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation.
  • Lecture 9 1 Interactive Proof Systems/Protocols

    Lecture 9 1 Interactive Proof Systems/Protocols

    CS 282A/MATH 209A: Foundations of Cryptography Prof. Rafail Ostrovsky Lecture 9 Lecture date: March 7-9, 2005 Scribe: S. Bhattacharyya, R. Deak, P. Mirzadeh 1 Interactive Proof Systems/Protocols 1.1 Introduction The traditional mathematical notion of a proof is a simple passive protocol in which a prover P outputs a complete proof to a verifier V who decides on its validity. The interaction in this traditional sense is minimal and one-way, prover → verifier. The observation has been made that allowing the verifier to interact with the prover can have advantages, for example proving the assertion faster or proving more expressive languages. This extension allows for the idea of interactive proof systems (protocols). The general framework of the interactive proof system (protocol) involves a prover P with an exponential amount of time (computationally unbounded) and a verifier V with a polyno- mial amount of time. Both P and V exchange multiple messages (challenges and responses), usually dependent upon outcomes of fair coin tosses which they may or may not share. It is easy to see that since V is a poly-time machine (PPT), only a polynomial number of messages may be exchanged between the two. P ’s objective is to convince (prove to) the verifier the truth of an assertion, e.g., claimed knowledge of a proof that x ∈ L. V either accepts or rejects the interaction with the P . 1.2 Definition of Interactive Proof Systems An interactive proof system for a language L is a protocol PV for communication between a computationally unbounded (exponential time) machine P and a probabilistic poly-time (PPT) machine V such that the protocol satisfies the properties of completeness and sound- ness.
  • The Correlation Among Software Complexity Metrics with Case Study

    The Correlation Among Software Complexity Metrics with Case Study

    International Journal of Advanced Computer Research (ISSN (print): 2249-7277 ISSN (online): 2277-7970) Volume-4 Number-2 Issue-15 June-2014 The Correlation among Software Complexity Metrics with Case Study Yahya Tashtoush1, Mohammed Al-Maolegi2, Bassam Arkok3 Abstract software product attributes such as functionality, quality, complexity, efficiency, reliability or People demand for software quality is growing maintainability. For example, a higher number of increasingly, thus different scales for the software code lines will lead to greater software complexity are growing fast to handle the quality of software. and so on. The software complexity metric is one of the measurements that use some of the internal The complexity of software effects on maintenance attributes or characteristics of software to know how activities like software testability, reusability, they effect on the software quality. In this paper, we understandability and modifiability. Software cover some of more efficient software complexity complexity is defined as ―the degree to which a metrics such as Cyclomatic complexity, line of code system or component has a design or implementation and Hallstead complexity metric. This paper that is difficult to understand and verify‖ [1]. All the presents their impacts on the software quality. It factors that make program difficult to understand are also discusses and analyzes the correlation between responsible for complexity. So it is necessary to find them. It finally reveals their relation with the measurements for software to reduce the impacts of number of errors using a real dataset as a case the complexity and guarantee the quality at the same study. time as much as possible.
  • Lecture 10: Learning DNF, AC0, Juntas Feb 15, 2007 Lecturer: Ryan O’Donnell Scribe: Elaine Shi

    Lecture 10: Learning DNF, AC0, Juntas Feb 15, 2007 Lecturer: Ryan O’Donnell Scribe: Elaine Shi

    Analysis of Boolean Functions (CMU 18-859S, Spring 2007) Lecture 10: Learning DNF, AC0, Juntas Feb 15, 2007 Lecturer: Ryan O’Donnell Scribe: Elaine Shi 1 Learning DNF in Almost Polynomial Time From previous lectures, we have learned that if a function f is ǫ-concentrated on some collection , then we can learn the function using membership queries in poly( , 1/ǫ)poly(n) log(1/δ) time.S |S| O( w ) In the last lecture, we showed that a DNF of width w is ǫ-concentrated on a set of size n ǫ , and O( w ) concluded that width-w DNFs are learnable in time n ǫ . Today, we shall improve this bound, by showing that a DNF of width w is ǫ-concentrated on O(w log 1 ) a collection of size w ǫ . We shall hence conclude that poly(n)-size DNFs are learnable in almost polynomial time. Recall that in the last lecture we introduced H˚astad’s Switching Lemma, and we showed that 1 DNFs of width w are ǫ-concentrated on degrees up to O(w log ǫ ). Theorem 1.1 (Hastad’s˚ Switching Lemma) Let f be computable by a width-w DNF, If (I, X) is a random restriction with -probability ρ, then d N, ∗ ∀ ∈ d Pr[DT-depth(fX→I) >d] (5ρw) I,X ≤ Theorem 1.2 If f is a width-w DNF, then f(U)2 ǫ ≤ |U|≥OX(w log 1 ) ǫ b O(w log 1 ) To show that a DNF of width w is ǫ-concentrated on a collection of size w ǫ , we also need the following theorem: Theorem 1.3 If f is a width-w DNF, then 1 |U| f(U) 2 20w | | ≤ XU b Proof: Let (I, X) be a random restriction with -probability 1 .