A New Approach to Proving Upper Bounds for MAX-2-SAT

Size: px
Start display at page:

Download "A New Approach to Proving Upper Bounds for MAX-2-SAT"

Transcription

1 A New Approach to Proving Upper Bounds for MAX-2-SAT Arist Kojevnikov Alexander S. Kulikov Abstract In this paper we present a new approach to proving upper bounds for the maximum 2-satisfiability problem (MAX-2-SAT). We present a new 2 K/5.5 -time algorithm for MAX-2-SAT, where K is the number of clauses in an input formula. We also obtain a 2 N/6 bound, where N is the number of variables in an input formula, for a particular case of MAX-2-SAT, where each variable appears in at most three 2-clauses. This immediately implies a 2 N/6 bound, where N is the number of vertices in an input graph, for the independent set problem on 3-regular graphs. The key point of our improvement is a combined complexity measure for estimating the running time of an algorithm. By using a new complexity measure we able to provide a much simpler proof of new upper bounds for MAX-2-SAT than proofs of previously known bounds. 1 Introduction 1.1 The MAX-SAT and MAX-2-SAT Problems. The maximum satisfiability problem (MAX- SAT) is: Given a formula in conjunctive normal form (CNF), find a maximal possible number of simultaneously satisfiable clauses of this formula. It is one of the most important optimization problems (MAX-SAT is a strengthening of the well-known satisfiability problem). Many practical and theoretical problems can be formulated in terms of MAX-SAT. For example, the maximum cut, vertex cover and maximum independent set problems can be reduced to MAX-2-SAT [14]. Also, MAX-SAT has applications in artificial intelligence and database systems [1, 6]. 1.2 Upper bounds. The decision version of MAX- SAT (where one is asked whether there exists an assignment to variables of a given formula satisfying at arist@logic.pdmi.ras.ru. St. Petersburg Department of Steklov Institute of Mathematics, 27 Fontanka, , St. Petersburg, Russia. Research supported in part by Russian Science Support Foundation, RFBR grant , and INTAS grant kulikov@logic.pdmi.ras.ru. St. Petersburg Department of Steklov Institute of Mathematics, 27 Fontanka, , St. Petersburg, Russia. Research supported in part by grant NSh of the President of Russian Federation for Leading Scientific School Support and INTAS grant Table 1: Known upper bounds for MAX-SAT and MAX-2-SAT MAX-SAT MAX-2-SAT N 2 N 2 N/1.261 [16] K 2 K/2.465 [3] 2 K/5.217 [9] L 2 L/6.892 [2] 2 L/ [7] least a given number of clauses) is NP-complete even in case, where each clause of an input formula contains at most two literals (this particular case of the MAX-SAT problem is called MAX-2-SAT). Moreover, it is known [15] that the decision version of MAX-2- SAT is NP-complete, if each variable appears in an input formula at most three times. In Table 1 we give the best known upper bounds for MAX-SAT and MAX-2- SAT w.r.t. the three standard formula complexity measures, namely the number of variables N, the number of clauses K, and the length L of an input formula. For simplicity, throughout all the paper we give exponential upper bounds without polynomial factors. Almost all algorithms from Table 1 (as well as our new algorithm) are based on a splitting method and thus use polynomial space. The only exception is the algorithm by Williams [16], which is based on different ideas (finding triangles in an auxiliar graph) and uses exponential space. The William s bound is the first improving the trivial 2 N bound. Recently, by proving a 2 N/6 bound for the pathwidth of cubic graphs Fomin et al. [5] obtained an exponential-space 2 N/6 -time algorithm for the independent set problem (and several other NP-hard graph problems). By using this result Kneis et al. [8] designed an exponential-space 2 K/ time algorithm for MAX- 2-SAT. 1.3 Our Result. In this paper we present a new 2 K/5.5 upper bound for MAX-2-SAT. The key point of our improvement is a combined complexity measure for estimating the running time of an algorithm. Let us explain informally what advantages one can get by using non-classical complexity measures (all formal definitions and proofs are given in the following sections).

2 Let (i, j)-literal denote a literal that occurs in a formula in i clauses positively and in j clauses negatively. Note that an (i, j)-literal provides an (i, j)-splitting w.r.t. the number of clauses K. Now let (i, j)-literal denote a literal occurring in i 2-clauses positively and in j 2-clauses negatively. Then an (i, j)-literal immediately gives an (i + j, i + j)-splitting w.r.t. the number of 2-clauses K 2. Moreover, if one proves an upper bound c K2 for MAX- 2-SAT, then a bound c K also holds, as K 2 K for any formula in 2-CNF. So, estimating the complexity of an algorithm not in terms of K, but in terms of K 2 is more convenient. Gramm et al. [7] use this idea to design a 2 K/5 -time algorithm for MAX-2-SAT. We also use a non-standard measure to prove a 2 K/5.5 upper bound for MAX-2-SAT. Our measure is obtained from K 2 by eliminating the dependency on variables that occur in at most two 2-clauses (such variables can be eliminated from a formula by simplification rules) and reducing the dependency on variables that occur in either three or four 2-clauses. It is known [14] that the maximum independent set problem can be easily reduced to MAX-2-SAT. For graph with n vertices and K edges, the reduction produces a 2-CNF formula having n variables and K 2-clauses. The maximum independent set problem is known to be NP-complete even for 3-regular graphs. Our 2 N/6 bound for the particular case of MAX-2- SAT where each variable appears in at most three 2- clauses in an input formula implies a 2 N/6 bound for the independent set problem on 3-regular graphs. 1.4 Automated Proofs of Upper Bounds. The algorithm of this paper is based on a splitting method. Currently best known upper bounds for many NP-hard problems are proven by using exactly this method. Usually the analysis of a splitting algorithm contains a big number of cases. This implies that many modern papers describing splitting algorithms are quite difficult to read and verify. For this reason, we also used a computer for analyzing the cases corresponding to the proof of the new bound. More precisely, we asked a computer to prove that the MAX-2-SAT problem can be solved in 2 N/6 time for formulas containing only variables occurring in at most three 2-clauses (the main case analysis of this paper is devoted to exactly this case). The automated proof of this fact is available at kulikov/autoproofs. The used framework for automated proofs of upper bounds on the running time of splitting algorithms is described in detail in [4]. This program uses the so-called general simplification rule [10], which generalizes such simplification rules as pure literal, dominating unit clause, almost dominating unit clause, rare variable. Note that the automated proof contains a list of bottleneck cases, which can be useful for further improving upper bounds for MAX-2-SAT. 2 General Setting 2.1 Basic Definitions. Let V be a set of Boolean variables. By v we denote the negation of a variable v V. A literal is either a Boolean variable or its negation. For a literal l, by var(l) we denote the variable corresponding to l. A clause is a disjunction of several literals l 1, l 2,..., l n, such that var(l i ) var(l j ) for i j. The length of a clause is the number of its literals. A k-clause is a clause consisting of exactly k literals. A formula in conjunctive normal form (CNF) is a conjunction of clauses. The length of a formula is the number of literals in it, that is, the sum of the lengths of all its clauses. A formula in k-cnf is a formula in CNF containing only clauses of length at most k. Following the notation of [7], in this paper we allow a formula to contain also a special true clause T, which is satisfied by any assignment. Informally speaking, the number of such clauses is the number of already satisfied clauses. By OptV al(f ) we denote the maximal number of simultaneously satisfiable clauses of F. We say that a literal l occurs in a clause, if this clause contains l. However, when we say that a variable x occurs in a clause, we mean that this clause contains either x or x. By #(F, l) we denote the number of occurrences of a literal l in a formula F, by # k (F, l) we mean the number of k-clauses of F containing l. By weight of a variable we mean the number of 2-clauses containing this variable (thus, if x is a variable of weight k, then # 2 (F, x) + # 2 (F, x) = k). By V (F ) we denote the set of all variables of F. Neighbors of a variable x are all variables occurring in 2-clauses with x. For a set of variables V 0 V (F ), by N(F, V 0 ) we denote the set of all neighbors of variables from V 0 except for the variables from V 0. By G(F ) we denote the following undirected multi-graph: V (F ) is the set of vertices, two vertices are connected by an edge if their corresponding variables occur in a 2-clause in F (note that this definition allows multiple edges between two vertices). We usually omit F when it is clear from the context. For a formula F and a literal l, by F [l] we denote a formula obtained from F by the following way: replace each clause of F containing l by T, eliminate all occurrences of l from other clauses, eliminate all empty clauses from F. By F [l 1 = l 2 ] we mean the following formula: replace all occurrences of l 1 by l 2 and all occurrences of l 1 by l 2, replace all clauses that contain a literal together with its negation by T, for each clause containing a pair of equal literals remove one of these

3 literals from this clause. 2.2 Complexity Measures. Usually to prove an upper bound on the running time of a splitting algorithm one fixes a formula complexity measure and shows that the algorithm always splits the formula into several formulas each having smaller complexity than the initial formula. The natural complexity measures are the number of variables N, the number of clauses K and the length L of a formula. A lot of bounds w.r.t. these measures for the SAT and MAX-SAT problems and their restricted versions are known. Sometimes however it is possible to use a different complexity measure to prove an upper bound w.r.t. one of the standard complexity measures. Such a trick is used, e.g., by Gramm et al. [7] to prove an upper bound 2 K/5 for MAX-2-SAT and by Kullmann [11] to prove an upper bound N for 3-SAT. In this paper we also use a non-standard complexity measure to prove a better upper bound for MAX-2-SAT w.r.t. K, its construction is given below. Let F denote a 2-CNF formula F. By N i (F ) we denote the number of variables of F occurring in exactly i 2-clauses in F. It is easy to see that the following equality holds: K 2 (F ) = N(F ) i=1 i N i (F ) 2 where K 2 (F ) is the number of 2-clauses in F. Our new complexity measure is obtained by a slight change of coefficients before the first four N i : N(F ) γ(f ) = N 3 (F ) N 4 (F ) + i=5, i N i (F ) 2 Let us explain (informally) how this change of coefficients helps to improve an upper bound for MAX-2-SAT. Note that splitting on a variable of weight w in general case gives a (w, w)-splitting w.r.t. K 2 (as in both branches all 2-clauses containing this variable are eliminated or become 1-clauses). Now consider a variable x of weight 5. If all neighbors of x are different and have weight at most 5, then splitting on x is a (5.5, 5.5)- splitting w.r.t. γ, as in both branches x is eliminated (γ is reduced by 2.5) and the weight of all its neighbors is decreased (γ is reduced by 0.6 5). Our analysis in the following sections shows that we can guarantee such a splitting in all other cases. 2.3 Estimating the Running Time. For estimating the running time of an algorithm we use the wellknown technique of recurrent inequalities [12]. For a formula F, a literal l and a complexity measure µ, we. say that a splitting F [l], F [ l] is an (i, j)-splitting w.r.t. µ, if µ(f ) µ(f [l]) i and µ(f ) µ(f [ l]) j. A splitting number of an (i, j)-splitting is the unique positive root of the equation x i + x j = 1, we denote it by τ(i, j). It is known that if an algorithm always splits with a splitting number not exceeding some constant c, then its running time on a formula F is bounded by c µ(f ). Our algorithm always performs (4, 10)- and (5.5, 5.5)-splitting w.r.t. γ. This implies that it works in time 2 γ/5.5 (as = τ(4, 10) < τ(5.5, 5.5) = 2 1/5.5 = ). 3 Simplification Rules We use the procedure Simp for simplifying a formula, see Fig. 1. It is easy to see that the running time of this procedure is polynomial in the length of an input formula. Its correctness as well as several additional properties are given in Lemma 3.1. We call a formula simplified if the procedure Simp does not change it. Lemma 3.1. Let F be a formula in 2-CNF and let F = Simp(F ). Then the following hold: 1. OptV al(f ) = OptV al(f ); 2. F does not contain variables of weight at most 2; 3. for any variable x, # 2 (F, x) # 2 (F, x); 4. γ(f ) γ(f ); 5. let F be a simplified formula, a, x be neighbors in F, and a has weight 3; then at least in one of the formulas F [x] and F [ x] the simplification rules assign a Boolean value to a. Proof. 1. Let F be a formula and l be a literal of F such that either #(F, l) = 0 or # 1 (F, l) #(F, l). Note that for any assignment A for all variables of the formula F, such that l A, an assignment A\{ l} {l} satisfies not less clauses of F than A. This shows that the pure literal and dominating unit clause rules preserve OptV al(f ). The correctness of the frequently meeting variable rule follows from the correctness of the pure literal and dominating unit clause rules. The almost common clauses rule is correct as any assignment satisfies at least one of the clauses C and D, and an assignment satisfies both of them iff it satisfies C\{l}. 2. F does not contain variables of weight at most 2, as all such variables can be eliminated by the frequently meeting variables rule.

4 Procedure Simp Input: A formula F in 2-CNF. Output: A simplified formula F in 2-CNF. Method. 1. Almost common clauses. If F = F 0 {C, D}, where C\{l} = D\{ l} for a literal l, then return Simp(F 0 {C\{x}, T }). 2. Pure literal. If #(F, l) = 0 for a literal l F, then return Simp(F [l]). 3. Dominating unit clause. If # 1 (F, l) #(F, l) for a literal l F, then return Simp(F [l]). 4. Frequently meeting variables. Let F contains two variables x 1 and x 2, such that x 1 occurs in F in at most one 2-clause without x 2. Note that after assigning x 2 a Boolean value, x 1 appears in at most one 2-clause, so either the pure literal or dominating unit clause rule assigns a Boolean value to x 1 (before that the almost common clauses rule possibly removes several unit clauses with x 1 ). Let α, β be Boolean values that the simplification rules assign to x 1 in F [x 2 ] and F [ x 2 ], respectively. Depending on the values of α and β the frequently meeting variables rule simplifies the formula F as follows: if α = 0, β = 0, then return Simp(F [ x 1 ]) if α = 0, β = 1, then return Simp(F [x 1 = x 2 ]) if α = 1, β = 0, then return Simp(F [x 1 = x 2 ]) if α = 1, β = 1, then return Simp(F [x 1 ]) 5. Return F. Figure 1: A procedure for simplifying a formula 3. It is easy to see that assigning a Boolean value to a variable as well as replacing two clauses by their common part do not increase the weight of variables. Let us show that the frequently meeting variables rule also does not increase the weight of variables. The only variable whose weight can possibly be increased is x 2. However, F [x 1 = x 2 ] (as well as F [x 1 = x 2 ]) does not contain 2-clauses of F consisting of x 1 and x 2 (at least one clause) and contains at most one new 2-clause (as x 1 appears without x 2 at most once). 4. Follows immediately from 3 and the definition of γ. 5. To prove this we need to consider several cases of other occurrences of the variable a. There are no other possible cases as F is simplified. (a) (ax)(a?)(ā?) after assigning x = 0 the dominating unit clause rule assigns a = 1; (b) (ax)(ā?)(ā?) after assigning x = 1 the pure literal rule assigns a = 0; (c) (ax)(a?)(a?)(ā) after assigning x = 0 the almost common clauses rule eliminates (a) and (ā), and then the pure literal rule assigns a = 1; (d) (ax)(a?)(a?)(ā)(ā) after assigning x = 1 the dominating until clause rule assigns a = 0; (e) (ax)(a?)(ā?)(ā) after assigning x = 1 the dominating unit clause rule assigns a = 0; (f) (ax)(ā?)(ā?)(a) after assigning x = 0 the dominating unit clause rule assigns a = 1. 4 Algorithm Our new algorithm is presented in Fig. 2. It is a typical splitting algorithm, that is, it first simplifies an input formula, then splits the simplified formula into two simpler formulas and recursively calls for them, and finally returns the answer according to the answers returned by both recursive calls. An upper bound on the running time of the algorithm is proven in Theorem 4.1. Theorem 4.1. Given a formula in 2-CNF, the algorithm MaxSatAlg returns the maximal possible number of simultaneously satisfiable clauses of this formula in time 2 K/5.5, where K is the number of clauses in the formula. Proof. The correctness of MaxSatAlg follows immediately from the correctness of the procedure Simp. To prove the stated upper bound we show that MaxSatAlg always splits with a splitting number not exceeding

5 Algorithm MaxSatAlg Input: A formula F in 2-CNF. Output: OptV al(f ). Method. 1. F = Simp(F ) 2. if F contains only (T ) clauses, then return L(F ) 3. if F = F 1 F 2, where F 1 and F 2 do not share variables, then return MaxSatAlg(F 1 ) + MaxSatAlg(F 2 ) 4. choose the variable x such that τ(γ(f ) γ(simp(f [x])), γ(f ) γ(simp(f [ x]))) is minimal and return max(m axsatalg(f [x]), M axsatalg(f [ x])) Figure 2: A 2 K/5.5 -Algorithm for MAX-2-SAT 2 1/5.5 w.r.t. γ. This is sufficient as γ(f ) K(F ) for any 2-CNF formula F. Consider a 2-CNF formula F, such that F is simplified and does not contain closed subformulas. Note that F does not contain variables of weight at most two. Below we consider several cases depending on the maximal weight of the variables of F. In each case we show that F contains a variable such that splitting on it is at least (5.5, 5.5)-splitting w.r.t. γ. Suppose that F contains a variable x of weight w 6. Observe that all variables of F (except for x) appear in at least two 2-clauses without x (otherwise, the frequently meeting variables rule is applicable). Also, reducing the weight of a variable of weight at least 3 reduces γ at least by 1/2 (as the difference between any two coefficients of N i in the definition of γ is at least 1/2). Thus, after assigning x a Boolean value, γ is reduced at least by w/2 (as x is eliminated) plus w/2 (as the weight of the neighbors of x is reduced). This means that splitting on x is at least (6, 6)-splitting w.r.t. γ. Now suppose that F contains only variables of weight at most 5. Let x be a variable of weight exactly 5. For 1 i j 5, let k ij denote the number of neighbors of x, that have weight j and occur i times with x. Since F is simplified, k ij = 0 for all i, j such that either j 2 (as the simplification rules eliminate variables of weight at most 2) or j i 1 (as the frequently meeting variables rule eliminates such variables). Thus, only the following k ij -s may have positive values: k 13, k 14, k 15, k 24, k 25, k 35. Since x occurs in five 2-clauses, the following equality holds: k 13 + k 14 + k k k k 35 = 5. Note that if a variable has weight j and occurs i times with x, after assigning x a Boolean value it occurs in (j i) 2-clauses. Let F be a formula obtained from F by assigning a Boolean value to x. Then, γ(f ) γ(f ) = (k (k 14 + k 24 ) + 2.5(k 15 + k 25 + k )) (k 14 + k k 15 ) 0.6(k 13 + k 14 + k k k k 35 ) = = 5.5. This shows that splitting on x is at least (5.5, 5.5)- splitting w.r.t. γ (as the procedure Simp does not increase γ). The case when F contains a variable of weight 4 (and no variables of higher weight) is similar. The condition that a variable occurs in four 2-clauses looks as follows: k 13 + k k 24 = 4, the difference of complexities is estimated as follows: γ(f ) γ(f ) = (k (k 14 + k )) (k 14 ) 0.9(k 13 + k k 24 ) = = 5.5. The only remaining case is when F contains only variables of weight 3. This case is handled in Theorem 4.2. Theorem 4.2. If a simplified formula F contains only variables of weight 3, then F contains a variable, such that splitting on it is at least (6, 6)-splitting w.r.t. γ. Proof. It is easy to see that γ(f ) = N(F ), so in the following we show that it is possible to find a (6, 6)- splitting w.r.t. the number of variables. Note that any formula consisting of variables of weight 3 only contains even number of variables. Also, if a formula is simplified, then it contains only variables of weight 3. Thus, each splitting in F is a (2k, 2l)-splitting for some integers k and l. Moreover, splitting on any variable of F immediately gives us a (4, 4)-splitting, since each variable has exactly three neighbors (as the frequently meeting variables rule is not applicable) and the simplification rules eliminate variable that occurs in less than three 2-clauses. In the following we show that F contains a variable x, such that splitting on it is either (4, 10)- or (6, 6)-splitting w.r.t. N. First, suppose that F contains three variables, such that any two of them appear in a 2-clause (such variables form a triangle in G(F )). Let a, b, c be the variables of this triangle. Observe that N({a, b, c}) contains at least two variables (otherwise, a, b, c and its neighbor form

6 least in one of the branches a, b, y, c, z are assigned Boolean values, all its neighbors are further eliminated). Figure 3: Main cases a closed subformula). At least on of these neighbors appears with only one variable from {a, b, c}, call it d and suppose w.l.o.g. that it occurs with a (see Fig. 3(a)). We state that splitting on d is a (6, 6)- splitting. After assigning d a Boolean value, the variable a appears only in 2-clauses with b and c and possibly in some unit clauses. The simplification rules either assign a Boolean value to a or replace both these 2-clauses by a clause consisting of variables b and c. In the former case both b and c are also eliminated by simplification rules. In the latter the frequently meeting variables rule eliminates one of the variables b and c. Thus, after assigning d a Boolean variable at least four variables are eliminated (three neighbors of d and at least one of b and c). This shows that splitting on d is a (6, 6)- splitting (as any splitting in F is a (2k, 2l)-splitting). Now consider the case when G(F ) does not contain triangles. Suppose that F contains a variable x with neighbors a, b, c, such that in both F [x] and F [ x] the simplification rules assign a Boolean value to at least one of the variables a, b, c. It is easy to see that splitting on x is a (6, 6)-splitting. Consider one of the branches F [x] and F [ x]. W.l.o.g. suppose that that a is assigned a Boolean value in this branch. Clearly, all the neighbors of a are eliminated. Since N({a}) {b, c} = (as G(F ) does not contain triangles), at least 4 variables are eliminated in this branch. Thus, in the following we suppose that after splitting on any variable of F the simplification rules assign Boolean values to all its neighbors in one branch (recall that due to Lemma 3.1 every neighbor is assigned a Boolean value in at least one branch). Consider a variable x with N({x}) = {a, b, c}. If N({a, b, c}) 5, then splitting on x is a (4, 10)-splitting. So, assume that N({a, b, c}) < 5. This means that either there is a variable y x, such that N(y) = {a, b, c} (see Fig. 3(b)), or there are two variables y, z x, such that N(y) {a, b, c} = 2 and N(z) {a, b, c} = 2 (see Fig. 3(c)). In the former case splitting on b is a (6, 6)- splitting (after eliminating x and y by the simplification rules, one of the variables a and c is also eliminated). In the latter case splitting on x is a (4, 10)-splitting (at 5 Conclusions and Further Directions We have presented the algorithm for MAX-2-SAT with the running time 2 K/5.5, where K is the number of clauses of an input formula. The key point of our improvement is a new complexity measure for estimating the running time of the algorithm. It would be interesting to design new measures to improve upper bounds for other NP-hard problems. It would also be interesting to apply the ideas from [13] to prove better upper bounds for MAX-2- SAT. Let us give several informal ideas showing how an 2 K/5 bound can be proven for MAX-2-SAT. Consider a splitting algorithm which uses the simplification rules of this paper, but splits on randomly chosen variables. Let us estimate for a variable x of weight w the probability that this algorithm will not split on x, that is, that x will be eliminated by the simplification rules. We know that the simplification rules eliminate variables of weight at most 2. Thus, in assumption that all neighbors of x are different, the required probability is at least 3/(w + 1), as it is exactly the probability that the algorithm will select x after all its neighbors except for at most two of them. Thus, the expected number of variables on which this algorithm splits is at most N(F ) i 2 i=1 i+1 N i(f ). It is easy to see that this sum does not exceed K 2 /5, i/10 for any positive integer i. Of course, this is only an informal proof, which needs to be done more accurate (for example, one has to consider cases when several neighbors of a variable are equal). We believe that this proof can be made precise. After doing this it is possible to estimate the probability that neighbors of a variable are eliminated not during splitting, but by as i 2 i+1 the simplification rules. This would allow to prove a stronger bound on the probability that the algorithm will not split on a variable. But such an estimation seems to be a more complicated work. Acknowledgments We would like to thank our supervisor Edward A. Hirsch for help in writing this paper. Also, we would like to thank the anonymous referees for helpful comments. References [1] P. Asirelli, M. De Santis, and M. Martelli. Integrity constraints in logic databases. Journal of Logic Programming, 3: , [2] N. Bansal and V. Raman. Upper bounds for MaxSat: Further improved. In Proceedings of ISAAC 99, pages , 1999.

7 [3] J. Chen and I. Kanj. Improved exact algorithms for MAX-SAT. In Proceedings of the 5th LATIN, volume 2286 of LNCS, pages , [4] S. S. Fedin and A. S. Kulikov. Automated proofs of upper bounds on the running time of splitting algorithms. Zapiski nauchnykh seminarov POMI, 316: , [5] F. V. Fomin and K. Høie. Pathwidth of cubic graphs and exact algorithms. Information Processing Letters, To appear. [6] H. Gallaire, J. Minker, and J.-M. Nicolas. Logic and databases: A deductive approach. ACM Computing Surveys, 16(2): , June [7] J. Gramm, E. A. Hirsch, R. Niedermeier, and P. Rossmanith. Worst-case upper bounds for MAX-2-SAT with an application to MAX-CUT. Discrete Applied Mathematics, 130(2): , [8] J. Kneis, D. MD. Mölle, S. Richter, and P. Rossmanith. Pathwidth of cubic graphs and exact algorithms. In Proceedings of the 31st International Workshop on Graph-Theoretic Concepts in Computer Science, To appear. [9] J. Kneis and P. Rossmanith. A new satisfiability algorithm with applications to Max-Cut. Technical Report AIB , Department of Computer Science, RWTH Aachen University, April [10] A. S. Kulikov. Automated generation of simplification rules for SAT and MAXSAT. In Proceedings of the Eighth International Conference on Theory and Applications of Satisfiability Testing (SAT 2005), volume 3569 of LNCS, pages Springer Verlag, [11] O. Kullmann. New methods for 3-SAT decision and worst-case analysis. Theoretical Computer Science, 223:1 72, [12] O. Kullmann and H. Luckhardt. Algorithms for SAT/TAUT decision based on various measures. Preprint, [13] R. Paturi, P. Pudlák, M. E. Saks, and F. Zane. An improved exponential-time algorithm for k-sat. In Proceedings of the 39th Annual Symposium on Foundations of Computer Science. (FOCS-98), pages , Palo Alto, CA, November [14] S. Poljak and Z. Tuza. Maximum cuts and largest bipartite subgraphs. DIMACS Series in Discrete Mathematics and Theoretical Computer Science, 20: , [15] V. Raman, B. Ravikumar, and S. Srinivasa Rao. A simplified NP-complete MAXSAT problem. Information Processing Letters, 65(1):1 6, [16] R. Williams. A new algorithm for optimal constraint satisfaction and its implications. In Proceedings of the 31st International Colloquium on Automata, Languages and Programming (ICALP), volume 3142 of LNCS, pages Springer Verlag, 2004.

A New Upper Bound for Max-2-SAT: A Graph-Theoretic Approach

A New Upper Bound for Max-2-SAT: A Graph-Theoretic Approach A New Upper Bound for Max-2-SAT: A Graph-Theoretic Approach Daniel Raible & Henning Fernau University of Trier, FB 4 Abteilung Informatik, 54286 Trier, Germany {raible,fernau}@informatik.uni-trier.de Abstract.

More information

Worst-case upper bounds for MAX-2-SAT with an application to MAX-CUT

Worst-case upper bounds for MAX-2-SAT with an application to MAX-CUT Discrete Applied Mathematics 130 (2003) 139 155 www.elsevier.com/locate/dam Worst-case upper bounds for MAX-2-SAT with an application to MAX-CUT Jens Gramm a;1;2, Edward A. Hirsch b;3, Rolf Niedermeier

More information

Deterministic Algorithms for k-sat Based on Covering Codes and Local Search

Deterministic Algorithms for k-sat Based on Covering Codes and Local Search Deterministic Algorithms for k-sat Based on Covering Codes and Local Search Evgeny Dantsin 1,, Andreas Goerdt 2, Edward A. Hirsch 3,, and Uwe Schöning 4 1 Steklov Institute of Mathematics, 27 Fontanka,

More information

Clause Shortening Combined with Pruning Yields a New Upper Bound for Deterministic SAT Algorithms

Clause Shortening Combined with Pruning Yields a New Upper Bound for Deterministic SAT Algorithms Clause Shortening Combined with Pruning Yields a New Upper Bound for Deterministic SAT Algorithms Evgeny Dantsin,EdwardA.Hirsch 2,, and Alexander Wolpert Roosevelt University, 430 S. Michigan Av., Chicago,

More information

Improved Exact Algorithms for Max-Sat

Improved Exact Algorithms for Max-Sat Improved Exact Algorithms for Max-Sat Jianer Chen Iyad A. Kanj Abstract In this paper we present improved exact and parameterized algorithms for the maximum satisfiability problem. In particular, we give

More information

Branching. Teppo Niinimäki. Helsinki October 14, 2011 Seminar: Exact Exponential Algorithms UNIVERSITY OF HELSINKI Department of Computer Science

Branching. Teppo Niinimäki. Helsinki October 14, 2011 Seminar: Exact Exponential Algorithms UNIVERSITY OF HELSINKI Department of Computer Science Branching Teppo Niinimäki Helsinki October 14, 2011 Seminar: Exact Exponential Algorithms UNIVERSITY OF HELSINKI Department of Computer Science 1 For a large number of important computational problems

More information

An Improved Upper Bound for SAT

An Improved Upper Bound for SAT An Improved Upper Bound for SAT Evgeny Dantsin and Alexander Wolpert Roosevelt University, 430 S. Michigan Av., Chicago, IL 60605, USA {edantsin, awolpert}@roosevelt.edu Abstract. We give a randomized

More information

An algorithm for the satisfiability problem of formulas in conjunctive normal form

An algorithm for the satisfiability problem of formulas in conjunctive normal form Journal of Algorithms 54 (2005) 40 44 www.elsevier.com/locate/jalgor An algorithm for the satisfiability problem of formulas in conjunctive normal form Rainer Schuler Abt. Theoretische Informatik, Universität

More information

Solving SAT for CNF formulas with a one-sided restriction on variable occurrences

Solving SAT for CNF formulas with a one-sided restriction on variable occurrences Solving SAT for CNF formulas with a one-sided restriction on variable occurrences Daniel Johannsen 1, Igor Razgon 2, and Magnus Wahlström 1 1 Max-Planck-Institut für Informatik, Saarbrücken, Germany 2

More information

Comp487/587 - Boolean Formulas

Comp487/587 - Boolean Formulas Comp487/587 - Boolean Formulas 1 Logic and SAT 1.1 What is a Boolean Formula Logic is a way through which we can analyze and reason about simple or complicated events. In particular, we are interested

More information

Worst-Case Upper Bound for (1, 2)-QSAT

Worst-Case Upper Bound for (1, 2)-QSAT Worst-Case Upper Bound for (1, 2)-QSAT Minghao Yin Department of Computer, Northeast Normal University, Changchun, China, 130117 ymh@nenu.edu.cn Abstract. The rigorous theoretical analysis of the algorithm

More information

Trichotomy Results on the Complexity of Reasoning with Disjunctive Logic Programs

Trichotomy Results on the Complexity of Reasoning with Disjunctive Logic Programs Trichotomy Results on the Complexity of Reasoning with Disjunctive Logic Programs Mirosław Truszczyński Department of Computer Science, University of Kentucky, Lexington, KY 40506, USA Abstract. We present

More information

On Quantum Versions of Record-Breaking Algorithms for SAT

On Quantum Versions of Record-Breaking Algorithms for SAT On Quantum Versions of Record-Breaking Algorithms for SAT Evgeny Dantsin 1, Vladik Kreinovich 2, and Alexander Wolpert 1 1 Department of Computer Science, Roosevelt University Chicago, IL 60605, USA, {edantsin,awolpert}@roosevelt.edu

More information

The Parameterized Complexity of k-flip Local Search for SAT and MAX SAT

The Parameterized Complexity of k-flip Local Search for SAT and MAX SAT The Parameterized Complexity of k-flip Local Search for SAT and MAX SAT Stefan Szeider Vienna University of Technology, Institute of Information Systems Favoritenstraße 9-11, A-1040 Vienna, Austria, stefan@szeider.net

More information

On the Structure and the Number of Prime Implicants of 2-CNFs

On the Structure and the Number of Prime Implicants of 2-CNFs On the Structure and the Number of Prime Implicants of 2-CNFs Navid Talebanfard Department of Mathematical and Computing Sciences, Tokyo Institute of Technology, Meguro-ku Ookayama 2-12-1, Japan 152-8552

More information

The Cook-Levin Theorem

The Cook-Levin Theorem An Exposition Sandip Sinha Anamay Chaturvedi Indian Institute of Science, Bangalore 14th November 14 Introduction Deciding a Language Let L {0, 1} be a language, and let M be a Turing machine. We say M

More information

A Probabilistic Algorithm for -SAT Based on Limited Local Search and Restart

A Probabilistic Algorithm for -SAT Based on Limited Local Search and Restart A Probabilistic Algorithm for -SAT Based on Limited Local Search and Restart Uwe Schöning Universität Ulm, Abteilung Theoretische Informatik James-Franck-Ring, D-89069 Ulm, Germany e-mail: schoenin@informatik.uni-ulm.de

More information

NP Completeness. CS 374: Algorithms & Models of Computation, Spring Lecture 23. November 19, 2015

NP Completeness. CS 374: Algorithms & Models of Computation, Spring Lecture 23. November 19, 2015 CS 374: Algorithms & Models of Computation, Spring 2015 NP Completeness Lecture 23 November 19, 2015 Chandra & Lenny (UIUC) CS374 1 Spring 2015 1 / 37 Part I NP-Completeness Chandra & Lenny (UIUC) CS374

More information

Lecture 9: The Splitting Method for SAT

Lecture 9: The Splitting Method for SAT Lecture 9: The Splitting Method for SAT 1 Importance of SAT Cook-Levin Theorem: SAT is NP-complete. The reason why SAT is an important problem can be summarized as below: 1. A natural NP-Complete problem.

More information

A New Lower Bound on the Maximum Number of Satisfied Clauses in Max-SAT and its Algorithmic Applications

A New Lower Bound on the Maximum Number of Satisfied Clauses in Max-SAT and its Algorithmic Applications A New Lower Bound on the Maximum Number of Satisfied Clauses in Max-SAT and its Algorithmic Applications Robert Crowston, Gregory Gutin, Mark Jones, Anders Yeo Department of Computer Science Royal Holloway,

More information

Solvers for the Problem of Boolean Satisfiability (SAT) Will Klieber Aug 31, 2011

Solvers for the Problem of Boolean Satisfiability (SAT) Will Klieber Aug 31, 2011 Solvers for the Problem of Boolean Satisfiability (SAT) Will Klieber 15-414 Aug 31, 2011 Why study SAT solvers? Many problems reduce to SAT. Formal verification CAD, VLSI Optimization AI, planning, automated

More information

UC Berkeley CS 170: Efficient Algorithms and Intractable Problems Handout 22 Lecturer: David Wagner April 24, Notes 22 for CS 170

UC Berkeley CS 170: Efficient Algorithms and Intractable Problems Handout 22 Lecturer: David Wagner April 24, Notes 22 for CS 170 UC Berkeley CS 170: Efficient Algorithms and Intractable Problems Handout 22 Lecturer: David Wagner April 24, 2003 Notes 22 for CS 170 1 NP-completeness of Circuit-SAT We will prove that the circuit satisfiability

More information

NP and Computational Intractability

NP and Computational Intractability NP and Computational Intractability 1 Polynomial-Time Reduction Desiderata'. Suppose we could solve X in polynomial-time. What else could we solve in polynomial time? don't confuse with reduces from Reduction.

More information

Knowledge base (KB) = set of sentences in a formal language Declarative approach to building an agent (or other system):

Knowledge base (KB) = set of sentences in a formal language Declarative approach to building an agent (or other system): Logic Knowledge-based agents Inference engine Knowledge base Domain-independent algorithms Domain-specific content Knowledge base (KB) = set of sentences in a formal language Declarative approach to building

More information

Part 1: Propositional Logic

Part 1: Propositional Logic Part 1: Propositional Logic Literature (also for first-order logic) Schöning: Logik für Informatiker, Spektrum Fitting: First-Order Logic and Automated Theorem Proving, Springer 1 Last time 1.1 Syntax

More information

A brief introduction to Logic. (slides from

A brief introduction to Logic. (slides from A brief introduction to Logic (slides from http://www.decision-procedures.org/) 1 A Brief Introduction to Logic - Outline Propositional Logic :Syntax Propositional Logic :Semantics Satisfiability and validity

More information

MATHEMATICAL ENGINEERING TECHNICAL REPORTS. Deductive Inference for the Interiors and Exteriors of Horn Theories

MATHEMATICAL ENGINEERING TECHNICAL REPORTS. Deductive Inference for the Interiors and Exteriors of Horn Theories MATHEMATICAL ENGINEERING TECHNICAL REPORTS Deductive Inference for the Interiors and Exteriors of Horn Theories Kazuhisa MAKINO and Hirotaka ONO METR 2007 06 February 2007 DEPARTMENT OF MATHEMATICAL INFORMATICS

More information

Parameterized Algorithms and Kernels for 3-Hitting Set with Parity Constraints

Parameterized Algorithms and Kernels for 3-Hitting Set with Parity Constraints Parameterized Algorithms and Kernels for 3-Hitting Set with Parity Constraints Vikram Kamat 1 and Neeldhara Misra 2 1 University of Warsaw vkamat@mimuw.edu.pl 2 Indian Institute of Science, Bangalore neeldhara@csa.iisc.ernet.in

More information

Critical Reading of Optimization Methods for Logical Inference [1]

Critical Reading of Optimization Methods for Logical Inference [1] Critical Reading of Optimization Methods for Logical Inference [1] Undergraduate Research Internship Department of Management Sciences Fall 2007 Supervisor: Dr. Miguel Anjos UNIVERSITY OF WATERLOO Rajesh

More information

Improved Edge-Coloring with Three Colors

Improved Edge-Coloring with Three Colors Improved Edge-Coloring with Three Colors Łukasz Kowalik 1,2 1 Institute of Informatics, Warsaw University, Banacha 2, 02-097, Warsaw, Poland 2 Max-Planck-Institute für Informatik, Stuhlsatzenhausweg 85,

More information

Loop Formulas for Disjunctive Logic Programs

Loop Formulas for Disjunctive Logic Programs Nineteenth International Conference on Logic Programming (ICLP-03), pages 451-465, Mumbai, India, 2003 Loop Formulas for Disjunctive Logic Programs Joohyung Lee and Vladimir Lifschitz Department of Computer

More information

Elementary Sets for Logic Programs

Elementary Sets for Logic Programs Elementary Sets for Logic Programs Martin Gebser Institut für Informatik Universität Potsdam, Germany Joohyung Lee Computer Science and Engineering Arizona State University, USA Yuliya Lierler Department

More information

Phase transitions in Boolean satisfiability and graph coloring

Phase transitions in Boolean satisfiability and graph coloring Phase transitions in Boolean satisfiability and graph coloring Alexander Tsiatas May 8, 2008 Abstract I analyzed the behavior of the known phase transitions in two NPcomplete problems, 3-colorability and

More information

Exact Max 2-SAT: Easier and Faster. Martin Fürer Shiva Prasad Kasiviswanathan Pennsylvania State University, U.S.A

Exact Max 2-SAT: Easier and Faster. Martin Fürer Shiva Prasad Kasiviswanathan Pennsylvania State University, U.S.A Exact Max 2-SAT: Easier and Faster Martin Fürer Shiva Prasad Kasiviswanathan Pennsylvania State University, U.S.A MAX 2-SAT Input: A 2-CNF fomula F with weights on clauses. Good assignment is one that

More information

A Duality between Clause Width and Clause Density for SAT

A Duality between Clause Width and Clause Density for SAT A Duality between Clause Width and Clause Density for SAT Chris Calabro, Russell Impagliazzo, Ramamohan Paturi University of California, San Diego 9500 Gilman Drive, La Jolla, CA 92093 ccalabro,russell,paturi

More information

Conjunctive Normal Form and SAT

Conjunctive Normal Form and SAT Notes on Satisfiability-Based Problem Solving Conjunctive Normal Form and SAT David Mitchell mitchell@cs.sfu.ca October 4, 2015 These notes are a preliminary draft. Please use freely, but do not re-distribute

More information

NP-Complete Reductions 1

NP-Complete Reductions 1 x x x 2 x 2 x 3 x 3 x 4 x 4 CS 4407 2 22 32 Algorithms 3 2 23 3 33 NP-Complete Reductions Prof. Gregory Provan Department of Computer Science University College Cork Lecture Outline x x x 2 x 2 x 3 x 3

More information

Elementary Sets for Logic Programs

Elementary Sets for Logic Programs Elementary Sets for Logic Programs Martin Gebser Institut für Informatik Universität Potsdam, Germany Joohyung Lee Computer Science and Engineering Arizona State University, USA Yuliya Lierler Department

More information

CMSC 858F: Algorithmic Lower Bounds Fall SAT and NP-Hardness

CMSC 858F: Algorithmic Lower Bounds Fall SAT and NP-Hardness CMSC 858F: Algorithmic Lower Bounds Fall 2014 3-SAT and NP-Hardness Instructor: Mohammad T. Hajiaghayi Scribe: Philip Dasler September 23, 2014 The most important NP-Complete (logic) problem family! 1

More information

P, NP, NP-Complete, and NPhard

P, NP, NP-Complete, and NPhard P, NP, NP-Complete, and NPhard Problems Zhenjiang Li 21/09/2011 Outline Algorithm time complicity P and NP problems NP-Complete and NP-Hard problems Algorithm time complicity Outline What is this course

More information

Measure and Conquer: Domination A Case Study

Measure and Conquer: Domination A Case Study Measure and Conquer: Domination A Case Study Fedor V. Fomin 1, Fabrizio Grandoni 2, and Dieter Kratsch 3 1 Department of Informatics, University of Bergen, N-5020 Bergen, Norway, fomin@ii.uib.no 2 Dipartimento

More information

NP-COMPLETE PROBLEMS. 1. Characterizing NP. Proof

NP-COMPLETE PROBLEMS. 1. Characterizing NP. Proof T-79.5103 / Autumn 2006 NP-complete problems 1 NP-COMPLETE PROBLEMS Characterizing NP Variants of satisfiability Graph-theoretic problems Coloring problems Sets and numbers Pseudopolynomial algorithms

More information

Complexity of locally injective homomorphism to the Theta graphs

Complexity of locally injective homomorphism to the Theta graphs Complexity of locally injective homomorphism to the Theta graphs Bernard Lidický and Marek Tesař Department of Applied Mathematics, Charles University, Malostranské nám. 25, 118 00 Prague, Czech Republic

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

Propositional logic. First order logic. Alexander Clark. Autumn 2014

Propositional logic. First order logic. Alexander Clark. Autumn 2014 Propositional logic First order logic Alexander Clark Autumn 2014 Formal Logic Logical arguments are valid because of their form. Formal languages are devised to express exactly that relevant form and

More information

Formal Verification Methods 1: Propositional Logic

Formal Verification Methods 1: Propositional Logic Formal Verification Methods 1: Propositional Logic John Harrison Intel Corporation Course overview Propositional logic A resurgence of interest Logic and circuits Normal forms The Davis-Putnam procedure

More information

Propositional Logic: Evaluating the Formulas

Propositional Logic: Evaluating the Formulas Institute for Formal Models and Verification Johannes Kepler University Linz VL Logik (LVA-Nr. 342208) Winter Semester 2015/2016 Propositional Logic: Evaluating the Formulas Version 2015.2 Armin Biere

More information

Generalizations of Matched CNF Formulas

Generalizations of Matched CNF Formulas Generalizations of Matched CNF Formulas Stefan Szeider (szeider@cs.toronto.edu) Department of Computer Science, University of Toronto, M5S 3G4 Toronto, Ontario, Canada Abstract. A CNF formula is called

More information

Chapter 2. Reductions and NP. 2.1 Reductions Continued The Satisfiability Problem (SAT) SAT 3SAT. CS 573: Algorithms, Fall 2013 August 29, 2013

Chapter 2. Reductions and NP. 2.1 Reductions Continued The Satisfiability Problem (SAT) SAT 3SAT. CS 573: Algorithms, Fall 2013 August 29, 2013 Chapter 2 Reductions and NP CS 573: Algorithms, Fall 2013 August 29, 2013 2.1 Reductions Continued 2.1.1 The Satisfiability Problem SAT 2.1.1.1 Propositional Formulas Definition 2.1.1. Consider a set of

More information

CMSC 858F: Algorithmic Lower Bounds Fall SAT and NP-Hardness

CMSC 858F: Algorithmic Lower Bounds Fall SAT and NP-Hardness CMSC 858F: Algorithmic Lower Bounds Fall 2014 3-SAT and NP-Hardness Instructor: Mohammad T. Hajiaghayi September 23, 2014 The most important NP-Complete (logic) problem family! 1 SAT = Satisfiability The

More information

SAT, NP, NP-Completeness

SAT, NP, NP-Completeness CS 473: Algorithms, Spring 2018 SAT, NP, NP-Completeness Lecture 22 April 13, 2018 Most slides are courtesy Prof. Chekuri Ruta (UIUC) CS473 1 Spring 2018 1 / 57 Part I Reductions Continued Ruta (UIUC)

More information

Design and Analysis of Algorithms

Design and Analysis of Algorithms Design and Analysis of Algorithms CSE 5311 Lecture 25 NP Completeness Junzhou Huang, Ph.D. Department of Computer Science and Engineering CSE5311 Design and Analysis of Algorithms 1 NP-Completeness Some

More information

Study of Lower Bound Functions for MAX-2-SAT

Study of Lower Bound Functions for MAX-2-SAT Study of Lower Bound Functions for MAX-2-SAT Haiou Shen, Hantao Zhang Computer Science Department The University of Iowa Iowa City, IA 52242 {hshen, hzhang}@cs.uiowa.edu Abstract Recently, several lower

More information

6.045: Automata, Computability, and Complexity (GITCS) Class 15 Nancy Lynch

6.045: Automata, Computability, and Complexity (GITCS) Class 15 Nancy Lynch 6.045: Automata, Computability, and Complexity (GITCS) Class 15 Nancy Lynch Today: More Complexity Theory Polynomial-time reducibility, NP-completeness, and the Satisfiability (SAT) problem Topics: Introduction

More information

Warm-Up Problem. Is the following true or false? 1/35

Warm-Up Problem. Is the following true or false? 1/35 Warm-Up Problem Is the following true or false? 1/35 Propositional Logic: Resolution Carmen Bruni Lecture 6 Based on work by J Buss, A Gao, L Kari, A Lubiw, B Bonakdarpour, D Maftuleac, C Roberts, R Trefler,

More information

Reasoning with Quantified Boolean Formulas

Reasoning with Quantified Boolean Formulas Reasoning with Quantified Boolean Formulas Martina Seidl Institute for Formal Models and Verification Johannes Kepler University Linz 1 What are QBF? Quantified Boolean formulas (QBF) are formulas of propositional

More information

Part V. Intractable Problems

Part V. Intractable Problems Part V Intractable Problems 507 Chapter 16 N P-Completeness Up to now, we have focused on developing efficient algorithms for solving problems. The word efficient is somewhat subjective, and the degree

More information

LOGIC PROPOSITIONAL REASONING

LOGIC PROPOSITIONAL REASONING LOGIC PROPOSITIONAL REASONING WS 2017/2018 (342.208) Armin Biere Martina Seidl biere@jku.at martina.seidl@jku.at Institute for Formal Models and Verification Johannes Kepler Universität Linz Version 2018.1

More information

Advanced Topics in Theoretical Computer Science

Advanced Topics in Theoretical Computer Science Advanced Topics in Theoretical Computer Science Part 5: Complexity (Part II) 30.01.2014 Viorica Sofronie-Stokkermans Universität Koblenz-Landau e-mail: sofronie@uni-koblenz.de 1 Contents Recall: Turing

More information

Aachen. A New Satisfiability Algorithm With Applications To Max-Cut. Department of Computer Science. Technical Report

Aachen. A New Satisfiability Algorithm With Applications To Max-Cut. Department of Computer Science. Technical Report Aachen Department of Computer Science Technical Report A New Satisfiability Algorithm With Applications To Max-Cut Joachim Kneis and Peter Rossmanith ISSN 0935 3232 Aachener Informatik Berichte AIB-2005-08

More information

P is the class of problems for which there are algorithms that solve the problem in time O(n k ) for some constant k.

P is the class of problems for which there are algorithms that solve the problem in time O(n k ) for some constant k. Complexity Theory Problems are divided into complexity classes. Informally: So far in this course, almost all algorithms had polynomial running time, i.e., on inputs of size n, worst-case running time

More information

SAT, Coloring, Hamiltonian Cycle, TSP

SAT, Coloring, Hamiltonian Cycle, TSP 1 SAT, Coloring, Hamiltonian Cycle, TSP Slides by Carl Kingsford Apr. 28, 2014 Sects. 8.2, 8.7, 8.5 2 Boolean Formulas Boolean Formulas: Variables: x 1, x 2, x 3 (can be either true or false) Terms: t

More information

On Computing k-cnf Formula Properties

On Computing k-cnf Formula Properties On Computing k-cnf Formula Properties Ryan Williams Computer Science Department, Carnegie Mellon University Pittsburgh, PA 15213 USA Abstract. The latest generation of SAT solvers (e.g. [11, 8]) generally

More information

an efficient procedure for the decision problem. We illustrate this phenomenon for the Satisfiability problem.

an efficient procedure for the decision problem. We illustrate this phenomenon for the Satisfiability problem. 1 More on NP In this set of lecture notes, we examine the class NP in more detail. We give a characterization of NP which justifies the guess and verify paradigm, and study the complexity of solving search

More information

Quantum Versions of k-csp Algorithms: a First Step Towards Quantum Algorithms for Interval-Related Constraint Satisfaction Problems

Quantum Versions of k-csp Algorithms: a First Step Towards Quantum Algorithms for Interval-Related Constraint Satisfaction Problems Quantum Versions of k-csp Algorithms: a First Step Towards Quantum Algorithms for Interval-Related Constraint Satisfaction Problems ABSTRACT Evgeny Dantsin Alexander Wolpert Department of Computer Science

More information

On the correlation of parity and small-depth circuits

On the correlation of parity and small-depth circuits Electronic Colloquium on Computational Complexity, Report No. 137 (2012) On the correlation of parity and small-depth circuits Johan Håstad KTH - Royal Institute of Technology November 1, 2012 Abstract

More information

Solving Connected Dominating Set Faster than 2 n

Solving Connected Dominating Set Faster than 2 n Solving Connected Dominating Set Faster than 2 n Fedor V. Fomin 1, Fabrizio Grandoni 2, and Dieter Kratsch 3 1 Department of Informatics, University of Bergen, N-5020 Bergen, Norway, fomin@ii.uib.no 2

More information

Intro to Theory of Computation

Intro to Theory of Computation Intro to Theory of Computation LECTURE 25 Last time Class NP Today Polynomial-time reductions Adam Smith; Sofya Raskhodnikova 4/18/2016 L25.1 The classes P and NP P is the class of languages decidable

More information

Computing Loops with at Most One External Support Rule for Disjunctive Logic Programs

Computing Loops with at Most One External Support Rule for Disjunctive Logic Programs Computing Loops with at Most One External Support Rule for Disjunctive Logic Programs Xiaoping Chen 1, Jianmin Ji 1, and Fangzhen Lin 2 1 School of Computer Science and Technology, University of Science

More information

Interleaved Alldifferent Constraints: CSP vs. SAT Approaches

Interleaved Alldifferent Constraints: CSP vs. SAT Approaches Interleaved Alldifferent Constraints: CSP vs. SAT Approaches Frédéric Lardeux 3, Eric Monfroy 1,2, and Frédéric Saubion 3 1 Universidad Técnica Federico Santa María, Valparaíso, Chile 2 LINA, Université

More information

CSCI 1590 Intro to Computational Complexity

CSCI 1590 Intro to Computational Complexity CSCI 1590 Intro to Computational Complexity NP-Complete Languages John E. Savage Brown University February 2, 2009 John E. Savage (Brown University) CSCI 1590 Intro to Computational Complexity February

More information

Hardness of Approximation

Hardness of Approximation Hardness of Approximation We have seen several methods to find approximation algorithms for NP-hard problems We have also seen a couple of examples where we could show lower bounds on the achievable approxmation

More information

NP-Completeness Part II

NP-Completeness Part II NP-Completeness Part II Please evaluate this course on Axess. Your comments really do make a difference. Announcements Problem Set 8 due tomorrow at 12:50PM sharp with one late day. Problem Set 9 out,

More information

NP Complete Problems. COMP 215 Lecture 20

NP Complete Problems. COMP 215 Lecture 20 NP Complete Problems COMP 215 Lecture 20 Complexity Theory Complexity theory is a research area unto itself. The central project is classifying problems as either tractable or intractable. Tractable Worst

More information

On the Gap Between ess(f) and cnf size(f) (Extended Abstract)

On the Gap Between ess(f) and cnf size(f) (Extended Abstract) On the Gap Between and (Extended Abstract) Lisa Hellerstein and Devorah Kletenik Polytechnic Institute of NYU, 6 Metrotech Center, Brooklyn, N.Y., 11201 Abstract Given a Boolean function f, denotes the

More information

Conjunctive Normal Form and SAT

Conjunctive Normal Form and SAT Notes on Satisfiability-Based Problem Solving Conjunctive Normal Form and SAT David Mitchell mitchell@cs.sfu.ca September 19, 2013 This is a preliminary draft of these notes. Please do not distribute without

More information

Conjunctive Normal Form and SAT

Conjunctive Normal Form and SAT Notes on Satisfiability-Based Problem Solving Conjunctive Normal Form and SAT David Mitchell mitchell@cs.sfu.ca September 10, 2014 These notes are a preliminary draft. Please use freely, but do not re-distribute

More information

On the Gap between the Complexity of SAT and Minimization for Certain Classes of Boolean Formulas

On the Gap between the Complexity of SAT and Minimization for Certain Classes of Boolean Formulas On the Gap between the Complexity of SAT and Minimization for Certain Classes of Boolean Formulas Ondřej Čepek and Štefan Gurský Abstract It is a wellknown fact that the satisfiability problem (SAT) for

More information

1 Propositional Logic

1 Propositional Logic CS 2800, Logic and Computation Propositional Logic Lectures Pete Manolios Version: 384 Spring 2011 1 Propositional Logic The study of logic was initiated by the ancient Greeks, who were concerned with

More information

Introduction to Solving Combinatorial Problems with SAT

Introduction to Solving Combinatorial Problems with SAT Introduction to Solving Combinatorial Problems with SAT Javier Larrosa December 19, 2014 Overview of the session Review of Propositional Logic The Conjunctive Normal Form (CNF) Modeling and solving combinatorial

More information

A Note on Perfect Partial Elimination

A Note on Perfect Partial Elimination A Note on Perfect Partial Elimination Matthijs Bomhoff, Walter Kern, and Georg Still University of Twente, Faculty of Electrical Engineering, Mathematics and Computer Science, P.O. Box 217, 7500 AE Enschede,

More information

Parametrizing Above Guaranteed Values: MaxSat. and MaxCut. Meena Mahajan and Venkatesh Raman. The Institute of Mathematical Sciences,

Parametrizing Above Guaranteed Values: MaxSat. and MaxCut. Meena Mahajan and Venkatesh Raman. The Institute of Mathematical Sciences, ECCC TR97-033 FTP: WWW: Email: ftp.eccc.uni-trier.de:/pub/eccc/ http://www.eccc.uni-trier.de/eccc/ ftpmail@ftp.eccc.uni-trier.de with subject help eccc Parametrizing Above Guaranteed Values: MaxSat and

More information

Non-Deterministic Time

Non-Deterministic Time Non-Deterministic Time Master Informatique 2016 1 Non-Deterministic Time Complexity Classes Reminder on DTM vs NDTM [Turing 1936] (q 0, x 0 ) (q 1, x 1 ) Deterministic (q n, x n ) Non-Deterministic (q

More information

Exponential Complexity of Satisfiability Testing for Linear-Size Boolean Formulas

Exponential Complexity of Satisfiability Testing for Linear-Size Boolean Formulas Exponential Complexity of Satisfiability Testing for Linear-Size Boolean Formulas Evgeny Dantsin and Alexander Wolpert Department of Computer Science, Roosevelt University 430 S. Michigan Av., Chicago,

More information

Polynomial-Time Reductions

Polynomial-Time Reductions Reductions 1 Polynomial-Time Reductions Classify Problems According to Computational Requirements Q. Which problems will we be able to solve in practice? A working definition. [von Neumann 1953, Godel

More information

arxiv:quant-ph/ v1 15 Apr 2005

arxiv:quant-ph/ v1 15 Apr 2005 Quantum walks on directed graphs Ashley Montanaro arxiv:quant-ph/0504116v1 15 Apr 2005 February 1, 2008 Abstract We consider the definition of quantum walks on directed graphs. Call a directed graph reversible

More information

NP-Complete Problems. More reductions

NP-Complete Problems. More reductions NP-Complete Problems More reductions Definitions P: problems that can be solved in polynomial time (typically in n, size of input) on a deterministic Turing machine Any normal computer simulates a DTM

More information

The complexity of acyclic subhypergraph problems

The complexity of acyclic subhypergraph problems The complexity of acyclic subhypergraph problems David Duris and Yann Strozecki Équipe de Logique Mathématique (FRE 3233) - Université Paris Diderot-Paris 7 {duris,strozecki}@logique.jussieu.fr Abstract.

More information

Computational Complexity and Intractability: An Introduction to the Theory of NP. Chapter 9

Computational Complexity and Intractability: An Introduction to the Theory of NP. Chapter 9 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

More information

Introduction to Complexity Theory

Introduction to Complexity Theory Introduction to Complexity Theory Read K & S Chapter 6. Most computational problems you will face your life are solvable (decidable). We have yet to address whether a problem is easy or hard. Complexity

More information

Minimization of Matched Formulas

Minimization of Matched Formulas WDS'11 Proceedings of Contributed Papers, Part I, 101 105, 2011. ISBN 978-80-7378-184-2 MATFYZPRESS Minimization of Matched Formulas Š. Gurský Charles University, Faculty of Mathematics and Physics, Prague,

More information

A Lower Bound of 2 n Conditional Jumps for Boolean Satisfiability on A Random Access Machine

A Lower Bound of 2 n Conditional Jumps for Boolean Satisfiability on A Random Access Machine A Lower Bound of 2 n Conditional Jumps for Boolean Satisfiability on A Random Access Machine Samuel C. Hsieh Computer Science Department, Ball State University July 3, 2014 Abstract We establish a lower

More information

A Collection of Problems in Propositional Logic

A Collection of Problems in Propositional Logic A Collection of Problems in Propositional Logic Hans Kleine Büning SS 2016 Problem 1: SAT (respectively SAT) Instance: A propositional formula α (for SAT in CNF). Question: Is α satisfiable? The problems

More information

More on NP and Reductions

More on NP and Reductions Indian Institute of Information Technology Design and Manufacturing, Kancheepuram Chennai 600 127, India An Autonomous Institute under MHRD, Govt of India http://www.iiitdm.ac.in COM 501 Advanced Data

More information

Resolution Lower Bounds for the Weak Pigeonhole Principle

Resolution Lower Bounds for the Weak Pigeonhole Principle Electronic Colloquium on Computational Complexity, Report No. 21 (2001) Resolution Lower Bounds for the Weak Pigeonhole Principle Ran Raz Weizmann Institute, and The Institute for Advanced Study ranraz@wisdom.weizmann.ac.il

More information

Artificial Intelligence Chapter 7: Logical Agents

Artificial Intelligence Chapter 7: Logical Agents Artificial Intelligence Chapter 7: Logical Agents Michael Scherger Department of Computer Science Kent State University February 20, 2006 AI: Chapter 7: Logical Agents 1 Contents Knowledge Based Agents

More information

A necessary and sufficient condition for the existence of a spanning tree with specified vertices having large degrees

A necessary and sufficient condition for the existence of a spanning tree with specified vertices having large degrees A necessary and sufficient condition for the existence of a spanning tree with specified vertices having large degrees Yoshimi Egawa Department of Mathematical Information Science, Tokyo University of

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

Computing branchwidth via efficient triangulations and blocks

Computing branchwidth via efficient triangulations and blocks Computing branchwidth via efficient triangulations and blocks Fedor Fomin Frédéric Mazoit Ioan Todinca Abstract Minimal triangulations and potential maximal cliques are the main ingredients for a number

More information

Detecting Backdoor Sets with Respect to Horn and Binary Clauses

Detecting Backdoor Sets with Respect to Horn and Binary Clauses Detecting Backdoor Sets with Respect to Horn and Binary Clauses Naomi Nishimura 1,, Prabhakar Ragde 1,, and Stefan Szeider 2, 1 School of Computer Science, University of Waterloo, Waterloo, Ontario, N2L

More information