Randomized Time Space Tradeoffs for Directed Graph Connectivity

Size: px
Start display at page:

Download "Randomized Time Space Tradeoffs for Directed Graph Connectivity"

Transcription

1 Randomized Time Space Tradeoffs for Directed Graph Connectivity Parikshit Gopalan, Richard Lipton, and Aranyak Mehta College of Computing, Georgia Institute of Technology, Atlanta GA 30309, USA. Abstract. We present a spectrum of randomized time-space tradeoffs for solving directed graph connectivity or STCONN in small space. We use a strategy parameterized by a parameter k that uses k pebbles and performs short random walks of length n k 1 using a probabilistic counter. We use this to get a family of algorithms that ranges between and log n in space and 2 log2 n and n n in running time. Our approach allows us to look at Savitch s algorithm and the random walk algorithm as two extremes of the same basic divide and conquer strategy. 1 Introduction The Graph Reachability problem is central to the study of algorithms that run in small space. Characterizing the complexities of the directed and undirected versions of this problem is an important problem that has received much attention. Undirected graph connectivity or USTCONN is in NL. It can also be decided in RL by taking a random walk on the graph [1]. Nisan [9] derandomizes the random walk using a pseudorandom generator for small space to prove that it lies in SC. Savitch s algorithm [13] is a deterministic algorithm which works in space. However it is also known to be in DSPACE(log 4 3 n) [2]. Feige [7] shows a family of randomized polynomial time algorithms that range between breadth first search and the RL random walk algorithm. Directed graph connectivity or STCONN is complete for NL under logspace reductions. It seems to have higher complexity than USTCONN for deterministic and randomized complexity classes. It is in DSPACE() by Savitch s theorem [13]. Savitch s algorithm is not a polynomial time algorithm, it runs in time 2 log2 n. There has been little success iesigning algorithms that simultaneously run in small space and time. STCONN is not known to lie in SC. The best tradeoff known currently [3] does give polynomial time for o(n) space, but not for space n 1 ε for any constant ε. It is also not known whether we can use Research supported in part by NSF CCR Also with Telcordia

2 randomness to reduce the space required. A simple random walk with restarts [8] works in space log n but the running time is doubly exponential in the space. Lower bounds for STCONN have been proved in several restricted models. See [12], [15] for a survey of some known results. In particular a lower bound of is known in the JAG model of Cook and Rackoff [5] and some of its extensions. For the RJAG model Berman and Simon [4] show that this bound holds for any algorithm running in time 2 logo(1) n. Poon [11] generalizes this result log to show a space lower bound of 2 n +log log T for an algorithm that runs in expected time T on a randomized NNJAG which generalizes the RJAG model. Our Results We present a spectrum of randomized time-space tradeoffs for solving directed graph connectivity or STCONN in small space. The spectrum ranges from Savitch s algorithm to the random walk with restarts algorithm of [8]. We use a strategy parameterized by a parameter k that uses k pebbles and performs short random walks of length d = n 1 k. Our main theorem is Theorem 1: There is a randomized algorithm that solves directed graph connectivity in RT ISP (2 d log2 n In particular, ) for 2 d n. When k = 1 and d = n this reduces to the random walk algorithm of [8] which solves STCONN in space log n and expected time n n. When k = log n and d = 2, we get a randomized analog of Savitch s algorithm [13] which takes space and time 2 log2 n. Intermediate values of k give an algorithm with parameters in between. If k = log n and time 2 log3 n. and d = log n, we get an algorithm that runs in space log 2 n Our algorithms can be made to run on the probabilistic NNJAG model defined by Poon in [11]. For space S log2 n, our algorithms exactly match the lower bounds shown by Poon. This shows that these lower bounds are optimal for a large range of parameters. Previously the only lower bounds known to be optimal in any JAG related model were for space Ω() due to Edmonds, Poon and Achlioptas [6]. Our algorithms use a divide and conquer strategy on the random walk. We divide a single random walk into a number of random walks of smaller size. This requires more space but speeds up the algorithm. It has been noted previously (e.g. [15]) that if we can solve s-t connectivity in log-space for any distance which is ω(1), then we can beat the space bound. That the random walk with restarts works in small space for any distance is also a classical result, as is the idea of small space probabilistic clocks [8], [14]. The main contribution of this

3 paper is to put these ideas together to yield an interesting trade-off result. The main lemmas involve a careful analysis of the time, randomness, space and error probability of our hierarchical random walks. Our approach allows us to look at Savitch s algorithm and the random walk algorithm as variants of the same basic divide and conquer strategy. In the next section we outline the idea behind the algorithm and analyze clocks that run in small space. The algorithm and its analysis are presented in full detail in Section 4. 2 Outline of the Algorithm Suppose we are given a directed graph G on n nodes and are asked to find a path of length d from u to v or report that no such path exists. One way to do this is by performing random walks of length d with restarts: Starting from u randomly choose the next step for d steps. If we reach v then we are done, else we restart at u. If indeed there is path between u and v of length d, we expect to find it in expected time. If we run for much longer (say time n 3d ) and do not find a path then we report that no such path exists and we will be correct with high probability. In [8] it is shown how this procedure can work in space log n. Our algorithm uses the above procedure of random walks with restarts. We use a divide and conquer strategy which is analogous to the recursive doubling strategy of Savitch s algorithm. Let us denote by G l the graph whose edges are paths in G of length l. If there is an s-t path in G, there is an s-t path of length at most n l in G l. Suppose we have k pebbles. Set d = n 1 k. We number our pebbles 1,..., k and use them in a hierarchical fashion. We try to discover an s-t path in G of length d. This we achieve by performing random walks of length d with restarts on G using pebble 1. However, we do not have G described explicitly, so we work recursively. We use pebble 2 to answer the queries is (x, y) an edge in G? ; and, in general, pebble i to answer edge queries about G n d i 1. The recursion bottoms out at pebble k, which we use to answer queries of the form is x connected to y by a path of length d in G?. Each level answers the queries asked by the previous level by performing random walks with restarts on the corresponding graph. A crucial requirement for performing all the random walks with restarts is a counter that keeps track of time up to n 3d so that we know when to stop the random walk and report that there is no u-v path. However, this is not possible in logspace. So we use a probabilistic counter (as in [8], [14]), which waits for a successive run of heads. We show that this counter works accurately enough to keep the errors small. An alternative way to represent this recursive pebbling strategy is to imagine that pebble 1 is performing a random walk on G, and that it queries an oracle for the edges of G. This oracle is implemented by pebbles 2 to k in a recursive manner.

4 Small Space Clocks Consider a sequence of unbiased coin tosses (0 or 1). Consider a counter which counts the lengths of a successive runs of 1s and stops at the first time it encounter a run of length l. We call this probabilistic counter Clock(2 l ). Lemma 1. Clock(2 l ) is a probabilistic procedure that runs in space log l for time T such that P r[t > t] < e t l2 l+1 and P r[t < t] < t 2 l Proof. Let T denote the time when a run of 1s of length l first appears. We first bound the probability that T is very large compared to 2 l. Divide the sequence into segments of length l. Let X i denote the random variable which is 1 if the i th segment contains all 1s, and is 0 otherwise. It is clear that the X i s are independent Bernoulli random variables each with expected value 2 l. Also, if T > t then X i = 0 for i = 1,..., t l. Therefore P r[t > t] P r[ t l i=1 X i = 0]. Hence, by the Chernoff bound, we get P r[t > t] < e t l2 l+1 (1) Next, a simple union bound shows that T is unlikely to be very small compared to 2 l. At any time, the probability that the next l coin tosses are all 1s is 2 l. Hence P r[t < t] < t 2 l (2) Note that the probability that the clock runs for a long time is much smaller than the probability that the clock runs out very quickly. However, this suffices for our purposes. 3 The Algorithm We describe a family of algorithms parameterized by a parameter k which takes values between 1 and log n. The input is a graph G on n vertices, and two vertices s and t. We set d = n 1 k. The algorithm is recursive. Each call is given input vertices u, v and a distance parameter d which is an upper bound on the distance from u to v. Thus the first call is with vertices s, t and d = n, and calls at the lowest level work with d = d. Each call performs a random walk with restarts on the appropriate graph and is clocked by Clock(n 3d ). As mentioned earlier, the first call is to RWALK(s, t, n). We first show that the algorithm indeed terminates with high probability. The main technical difficulty here is the following. The running time of the algorithm is essentially the sum of the running times of all the random walks performed. Each random walk runs for as long as the associated Clock runs. If this clock runs too long, so does the random walk. Each step of the random walk invokes a new random walk at the lower level (i.e. for a smaller distance). Hence

5 RWALK(u,v,d ) Start Clock(n 3d ) While (Clock(n 3d ) is ticking) v 0 = u; v cur = u; [RESTART] for i = 1 to d v prev = v cur R v cur [n] if d > d //(recursive call) if RWALK(v prev, v cur, d ) = 0 then goto [RESTART] d else //(bottom of recursion) if (v prev, v cur) / E(G) then goto [RESTART] end for if v cur = v then Return 1 else goto [RESTART] end while Return 0 //(Clock terminated) Fig. 1. RWALK(u,v,d ) there is no a priori bound even on the total number of random walks performed, it depends on the length of each random walk. We prove our bound through induction. By level j of the algorithm we will mean the jth level of recursion in the algorithm. Random walks at level j are performed using pebble j. There are k levels in all. Lemma 2. The algorithm terminates in time n O(dk) with very high probability. Proof. We use induction on the level j to prove the following: With probability at least 1 n7dj, for all l j, there are no more than n 7d(l 1) random walks e performed at the nd lth level of the recursion, and each of these random walks runs for time less than n 7d. The base case of the induction is true since only one random walk is performed at the first level and by Lemma 1 with probability at least 1 e nd it does not run for time greater than n 7d. Assume that the inductive hypothesis is true for level j. The number of random walks at level j + 1 is equal to the total number of steps performed in all the random walks at level j. Thus with probability at least 1 n7dj no more than n 7d(j 1) n 7d = n 7dj walks run at level j + 1. Given e nd this, the probability that any of these walks runs for more than n 7d is bounded by n7dj e nd. Hence the statement is true for j + 1 and this concludes the induction. Since there are k levels, with probability 1 n7dk e nd is bounded by k, the total number of walks i=1 n7d(i 1) n 7dk and none of them runs for more than n 7d. This implies that the algorithm runs in time n 7d(k+1) = n O(dk). We now argue that with good probability the algorithm does indeed find an s-t path if one exists.

6 Notice first that the algorithm only errs in one direction: if there is no s-t path the algorithm will certainly return 0. Now a call to RWALK (say RWALK(u, v, d )) at any level of the recursion essentially consists of two parts: while the clock runs it randomly chooses d vertices to perform the random walk between u and v; and it asks edge queries to the level below. Call a sequence of vertices x 0 = u, x 1, x 2,..., x d = v a valid sequence if in the graph G there is a path of length d /d from x i to x i+1 for all i = 0,..., d 1. We prove first that at any call RWALK(u, v, d ) finds a valid sequence with high probability if there is a path of length d from u to v in G. Lemma 3. If there is a u-v path of length d in G, RWALK(u, v, d ) finds a valid sequence with probability at least 1 2. Proof. By Lemma 1, we know that Clock(n 3d ) runs for time at least n 2d with probability at least 1 1. The probability that RWALK(u, v, d ) does not find a particular d length sequence in this time is at most 1. So the error probability is at most 2. Having RWALK(u, v, d ) choose a valid sequence with high probability is not enough. We also need the calls at all lower levels to return correct answers. Here the main issue is that we are running about k number of random walks, and the probability of error of each random walk is about n d. Hence clearly many of the walks will return incorrect answers. But the key is that for most instances, we do not care. We only require those queries which actually pertain to the s-t path to be answered correctly. For this we argue in a bottom up manner. For convenience, we define a notion of height of a pebble which is the inverse of the level. The pebble at level k is at height 1 whereas the pebble at level 1 is at height k. Lemma 4. If there is a s-t path in G, RWALK(s, t, n) will find it with probability at least Proof. By induction on h, we prove that the pebble at height h answers any single query incorrectly with probability at most h 1 i=0 2di. If h = 1, then this follows from Lemma 3 since finding a path between u and v is equivalent to finding a valid sequence (since at height 1 we have direct access to G, and do not need to make any edge queries). Assume that the statement holds for height h 1. Consider the pebble at height h which is asked if u is adjacent to v. If a u-v path exists, then by Lemma 3, it guesses a valid sequence with probability at least 1 2. It now makes d queries to the pebble at level h 1. That pebble returns correct answers to all these d queries with probability at least 1 d h 2 i=0 2di. Hence the probability that pebble at height h answers the query incorrectly is at most d h 2 i=0 2di + 2 = h 1 i=0 2di. The overall failure probability of RWALK(s, t, n) is the probability that the pebble at height k answers the question is there a path from s to t? incorrectly. By the induction above, this is at most 2dk = 2n which goes to 0 asymptotically.

7 We finally prove: Lemma 5. The algorithm runs in space k log n = log2 n Proof. There are k levels of recursion. At each level there is only one random walk running at any given time. Each level keeps a clock, a start and end vertex and two current vertices. So each level takes O(log n) storage. Hence the total storage is k log n. We have proved our main Theorem: Theorem 1. There is an algorithm that solves directed graph connectivity in RT ISP (2 d log2 n, ) for 2 d n or equivalently in RT ISP 1 (2n k, k log n) for 1 k log n. It is interesting to note that by setting d = 2, we perform random walks of length 2. This is equivalent to guessing the midpoint v between s and t and then the midpoint of each of these paths and so on. If instead of guessing, we try all possible vertices, this is precisely Savitch s algorithm. Whe = n we get the logspace random walk with restarts of [8]. Intermediate values of d give an algorithm with parameters in between. In particular, if d = log n, we get an algorithm that runs in space log2 n and time n. 2log3 4 Implementation on a Probabilistic NNJAG The NNJAG model is a restricted model of computation for directed graph connectivity. It consists of an automaton with a set of internal states q, a set of pebbles k, and a transition function δ. A probabilistic NNJAG has access to random bits at each step and the transition function caepend on the bits read. At any time the NNJAG only knows the names of the nodes on which it has pebbles. It is allowed to move a pebble to an adjacent node, or to jump a pebble to a node on which there are already other pebbles. The space S is defined as k log n + log q. The time taken T is the number of moves made. We shall briefly give an informal description of how algorithm RWALK can be simulated on a probabilistic NNJAG [10]. The counter for the clock is maintained in the internal state of the NNJAG. Similarly at each step, the guess for the k vertices on the path is stored with the internal state. We then move pebbles along the edge and compare them with the nodes stored in the internal state to see if we have reached the target node. Theorem 2. [11] Any probabilistic NNJAG for STCONN on n node graphs that runs in expected time T and uses space S satisfies S log n +log log T. Theorem 3. For S, there exists a probabilistic NNJAG that meets the lower bound of Theorem 2.

8 Proof: Since the lower bound on S by Theorem 2 is less than or equal to, we can hope to show our algorithm is tight only when the space is less than log2 n or equivalently whe = Ω(log n). In Theorem 1, T = 2 d log2 n log log T = Θ( + ) = Θ() The bound implied by Theorem 2 for an NNJAG with this expected running time is ( log 2 ) n S = Ω + log log T ( log 2 ) n = Ω But this is exactly the space bound achieved by Theorem 1. Acknowledgments We thank C.K. Poon for discussions regarding the NNJAG model. We thank H. Venkateswaran for many helpful discussions. References 1. R. Aleliunas, R. Karp, R. Lipton, L. Lovasz, and C. Rakoff. Random walks, universal traversal sequences, and the complexity of maze problems. In 20th Annual Symposium on Foundations of Computer Science, pages , R. Armoni, A. Ta-Shma, A. Wigderson, and S. Zhou. An log(n) 4 3 space algorithm for s-t connectivity in undirected graphs. Journal of the ACM, 47: , G. Barnes, J. Buss, W. Ruzzo, and B. Schieber. A sublinear space, polynomial time algorithm for directed s-t connectivity. In 7th annual conference on Structure in Complexity Theory, pages 27 33, P. Berman and J. Simon. Lower bounds on graph threading by probabilistic machines. In 24th Annual Symposium on Foundations of Computer Science, pages , S. Cook and C. Rackoff. Space lower bounds for maze threadability on restricted machines. SIAM Journal on Computing, 9(3): , J. Edmonds, C. Poon, and D. Achlioptas. Tight lower bounds for st-connectivity on the NNJAG model. SIAM J. Comput., 28(6): , U. Feige. A spectrum of time-space tradeoffs for undirected s-t connectivity. Journal of Computer and System Sciences, 54(2): , J. Gill. Computational complexity of probabilistic turing machines. In Proc. 6th Annual ACM Symp. Theory of Computing, pages 91 95, N. Nisan. RL SC. Journal of Computational Complexity, 4, pages 1 11, C. Poon. Personal communication.

9 11. C. Poon. Space bounds for graph connectivity problems on node named jags and node ordered jags. In 34th Annual Symposium on Foundations of Computer Science, pages , C. Poon. On the complexity of the s-t connectivity problem. Ph.D Thesis, University of Toronto, W. Savitch. Relationships between nondeterministic and deterministic tape complexities. Journal of Computer and System Sciences, 4(2): , J. Simon. Space-bounded probabilistic turing machine complexity classes are closed under complement. In Proc. 13th Annual ACM Symp. Theory of Computing, pages , A. Wigderson. The complexity of graph connectivity. In Proceedings of the 17th Mathematical Foundations of Computer Science, pages , 1992.

U.C. Berkeley CS278: Computational Complexity Professor Luca Trevisan 9/6/2004. Notes for Lecture 3

U.C. Berkeley CS278: Computational Complexity Professor Luca Trevisan 9/6/2004. Notes for Lecture 3 U.C. Berkeley CS278: Computational Complexity Handout N3 Professor Luca Trevisan 9/6/2004 Notes for Lecture 3 Revised 10/6/04 1 Space-Bounded Complexity Classes A machine solves a problem using space s(

More information

Notes on Space-Bounded Complexity

Notes on Space-Bounded Complexity U.C. Berkeley CS172: Automata, Computability and Complexity Handout 6 Professor Luca Trevisan 4/13/2004 Notes on Space-Bounded Complexity These are notes for CS278, Computational Complexity, scribed by

More information

Notes on Space-Bounded Complexity

Notes on Space-Bounded Complexity U.C. Berkeley CS172: Automata, Computability and Complexity Handout 7 Professor Luca Trevisan April 14, 2015 Notes on Space-Bounded Complexity These are notes for CS278, Computational Complexity, scribed

More information

Universal Traversal Sequences for Expander Graphs

Universal Traversal Sequences for Expander Graphs Universal Traversal Sequences for Expander Graphs Shlomo Hoory Avi Wigderson Hebrew University, Jerusalem Keywords: Universal traversal sequence, space-bounded computation, explicit construction, connectivity,

More information

Advanced topic: Space complexity

Advanced topic: Space complexity Advanced topic: Space complexity CSCI 3130 Formal Languages and Automata Theory Siu On CHAN Chinese University of Hong Kong Fall 2016 1/28 Review: time complexity We have looked at how long it takes to

More information

Umans Complexity Theory Lectures

Umans Complexity Theory Lectures Umans Complexity Theory Lectures Lecture 8: Introduction to Randomized Complexity: - Randomized complexity classes, - Error reduction, - in P/poly - Reingold s Undirected Graph Reachability in RL Randomized

More information

Undirected ST-Connectivity in Log-Space. Omer Reingold. Presented by: Thang N. Dinh CISE, University of Florida, Fall, 2010

Undirected ST-Connectivity in Log-Space. Omer Reingold. Presented by: Thang N. Dinh CISE, University of Florida, Fall, 2010 Undirected ST-Connectivity in Log-Space Omer Reingold Presented by: Thang N. Dinh CISE, University of Florida, Fall, 2010 Undirected s-t connectivity(ustcon) Is t reachable from s in a undirected graph

More information

CSC 5170: Theory of Computational Complexity Lecture 4 The Chinese University of Hong Kong 1 February 2010

CSC 5170: Theory of Computational Complexity Lecture 4 The Chinese University of Hong Kong 1 February 2010 CSC 5170: Theory of Computational Complexity Lecture 4 The Chinese University of Hong Kong 1 February 2010 Computational complexity studies the amount of resources necessary to perform given computations.

More information

Space is a computation resource. Unlike time it can be reused. Computational Complexity, by Fu Yuxi Space Complexity 1 / 44

Space is a computation resource. Unlike time it can be reused. Computational Complexity, by Fu Yuxi Space Complexity 1 / 44 Space Complexity Space is a computation resource. Unlike time it can be reused. Computational Complexity, by Fu Yuxi Space Complexity 1 / 44 Synopsis 1. Space Bounded Computation 2. Logspace Reduction

More information

The Complexity of Graph Connectivity

The Complexity of Graph Connectivity The Complexity of Graph Connectivity Avi Wigderson Hebrew University and Princeton University February 11, 2003 Abstract In this paper we survey the major developments in understanding the complexity of

More information

Universal Traversal Sequences with Backtracking

Universal Traversal Sequences with Backtracking Universal Traversal Sequences with Backtracking Michal Koucký Department of Computer Science Rutgers University, Piscataway NJ 08854, USA mkoucky@paul.rutgers.edu April 9, 001 Abstract In this paper we

More information

6.045 Final Exam Solutions

6.045 Final Exam Solutions 6.045J/18.400J: Automata, Computability and Complexity Prof. Nancy Lynch, Nati Srebro 6.045 Final Exam Solutions May 18, 2004 Susan Hohenberger Name: Please write your name on each page. This exam is open

More information

Complexity Theory 112. Space Complexity

Complexity Theory 112. Space Complexity Complexity Theory 112 Space Complexity We ve already seen the definition SPACE(f(n)): the languages accepted by a machine which uses O(f(n)) tape cells on inputs of length n. Counting only work space NSPACE(f(n))

More information

PRGs for space-bounded computation: INW, Nisan

PRGs for space-bounded computation: INW, Nisan 0368-4283: Space-Bounded Computation 15/5/2018 Lecture 9 PRGs for space-bounded computation: INW, Nisan Amnon Ta-Shma and Dean Doron 1 PRGs Definition 1. Let C be a collection of functions C : Σ n {0,

More information

Outline. Complexity Theory. Example. Sketch of a log-space TM for palindromes. Log-space computations. Example VU , SS 2018

Outline. Complexity Theory. Example. Sketch of a log-space TM for palindromes. Log-space computations. Example VU , SS 2018 Complexity Theory Complexity Theory Outline Complexity Theory VU 181.142, SS 2018 3. Logarithmic Space Reinhard Pichler Institute of Logic and Computation DBAI Group TU Wien 3. Logarithmic Space 3.1 Computational

More information

CSE200: Computability and complexity Space Complexity

CSE200: Computability and complexity Space Complexity CSE200: Computability and complexity Space Complexity Shachar Lovett January 29, 2018 1 Space complexity We would like to discuss languages that may be determined in sub-linear space. Lets first recall

More information

Lecture 3. 1 Terminology. 2 Non-Deterministic Space Complexity. Notes on Complexity Theory: Fall 2005 Last updated: September, 2005.

Lecture 3. 1 Terminology. 2 Non-Deterministic Space Complexity. Notes on Complexity Theory: Fall 2005 Last updated: September, 2005. Notes on Complexity Theory: Fall 2005 Last updated: September, 2005 Jonathan Katz Lecture 3 1 Terminology For any complexity class C, we define the class coc as follows: coc def = { L L C }. One class

More information

20.1 2SAT. CS125 Lecture 20 Fall 2016

20.1 2SAT. CS125 Lecture 20 Fall 2016 CS125 Lecture 20 Fall 2016 20.1 2SAT We show yet another possible way to solve the 2SAT problem. Recall that the input to 2SAT is a logical expression that is the conunction (AND) of a set of clauses,

More information

IITM-CS6845: Theory Toolkit February 3, 2012

IITM-CS6845: Theory Toolkit February 3, 2012 IITM-CS6845: Theory Toolkit February 3, 2012 Lecture 4 : Derandomizing the logspace algorithm for s-t connectivity Lecturer: N S Narayanaswamy Scribe: Mrinal Kumar Lecture Plan:In this lecture, we will

More information

The space complexity of a standard Turing machine. The space complexity of a nondeterministic Turing machine

The space complexity of a standard Turing machine. The space complexity of a nondeterministic Turing machine 298 8. Space Complexity The space complexity of a standard Turing machine M = (Q,,,, q 0, accept, reject) on input w is space M (w) = max{ uav : q 0 w M u q av, q Q, u, a, v * } The space complexity of

More information

MTAT Complexity Theory October 13th-14th, Lecture 6

MTAT Complexity Theory October 13th-14th, Lecture 6 MTAT.07.004 Complexity Theory October 13th-14th, 2011 Lecturer: Peeter Laud Lecture 6 Scribe(s): Riivo Talviste 1 Logarithmic memory Turing machines working in logarithmic space become interesting when

More information

Notes on Complexity Theory Last updated: December, Lecture 27

Notes on Complexity Theory Last updated: December, Lecture 27 Notes on Complexity Theory Last updated: December, 2011 Jonathan Katz Lecture 27 1 Space-Bounded Derandomization We now discuss derandomization of space-bounded algorithms. Here non-trivial results can

More information

Downloaded 07/22/13 to Redistribution subject to SIAM license or copyright; see

Downloaded 07/22/13 to Redistribution subject to SIAM license or copyright; see SIAM J. COMPUT. Vol. 28, No. 3, pp. 1051 1072 c 1999 Society for Industrial and Applied Mathematics A TIME-SPACE TRADEOFF FOR UNDIRECTED GRAPH TRAVERSAL BY WALKING AUTOMATA PAUL BEAME, ALLAN BORODIN, PRABHAKAR

More information

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

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

More information

U.C. Berkeley CS278: Computational Complexity Professor Luca Trevisan August 30, Notes for Lecture 1

U.C. Berkeley CS278: Computational Complexity Professor Luca Trevisan August 30, Notes for Lecture 1 U.C. Berkeley CS278: Computational Complexity Handout N1 Professor Luca Trevisan August 30, 2004 Notes for Lecture 1 This course assumes CS170, or equivalent, as a prerequisite. We will assume that the

More information

ReachFewL = ReachUL. Department of Computer Science and Engineering University of Nebraska Lincoln

ReachFewL = ReachUL. Department of Computer Science and Engineering University of Nebraska Lincoln Electronic Colloquium on Computational Complexity, Report No. 60 (2011) ReachFewL = ReachUL Brady Garvin Derrick Stolee Raghunath Tewari N. V. Vinodchandran Department of Computer Science and Engineering

More information

CS151 Complexity Theory. Lecture 4 April 12, 2017

CS151 Complexity Theory. Lecture 4 April 12, 2017 CS151 Complexity Theory Lecture 4 A puzzle A puzzle: two kinds of trees depth n...... cover up nodes with c colors promise: never color arrow same as blank determine which kind of tree in poly(n, c) steps?

More information

Time-Space Trade-Os. For Undirected ST -Connectivity. on a JAG

Time-Space Trade-Os. For Undirected ST -Connectivity. on a JAG Time-Space Trade-Os For Undirected ST -Connectivity on a JAG Abstract Undirected st-connectivity is an important problem in computing. There are algorithms for this problem that use O (n) time and ones

More information

Min/Max-Poly Weighting Schemes and the NL vs UL Problem

Min/Max-Poly Weighting Schemes and the NL vs UL Problem Min/Max-Poly Weighting Schemes and the NL vs UL Problem Anant Dhayal Jayalal Sarma Saurabh Sawlani May 3, 2016 Abstract For a graph G(V, E) ( V = n) and a vertex s V, a weighting scheme (w : E N) is called

More information

CS278: Computational Complexity Spring Luca Trevisan

CS278: Computational Complexity Spring Luca Trevisan CS278: Computational Complexity Spring 2001 Luca Trevisan These are scribed notes from a graduate course on Computational Complexity offered at the University of California at Berkeley in the Spring of

More information

COL 352 Introduction to Automata and Theory of Computation Major Exam, Sem II , Max 80, Time 2 hr. Name Entry No. Group

COL 352 Introduction to Automata and Theory of Computation Major Exam, Sem II , Max 80, Time 2 hr. Name Entry No. Group COL 352 Introduction to Automata and Theory of Computation Major Exam, Sem II 2015-16, Max 80, Time 2 hr Name Entry No. Group Note (i) Write your answers neatly and precisely in the space provided with

More information

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

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

More information

Umans Complexity Theory Lectures

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

More information

Pseudorandom Generators for Regular Branching Programs

Pseudorandom Generators for Regular Branching Programs Pseudorandom Generators for Regular Branching Programs Mark Braverman Anup Rao Ran Raz Amir Yehudayoff Abstract We give new pseudorandom generators for regular read-once branching programs of small width.

More information

Apropos of an errata in ÜB 10 exercise 3

Apropos of an errata in ÜB 10 exercise 3 Apropos of an errata in ÜB 10 exercise 3 Komplexität von Algorithmen SS13 The last exercise of the last exercise sheet was incorrectly formulated and could not be properly solved. Since no one spotted

More information

Lecture 6: Oracle TMs, Diagonalization Limits, Space Complexity

Lecture 6: Oracle TMs, Diagonalization Limits, Space Complexity CSE 531: Computational Complexity I Winter 2016 Lecture 6: Oracle TMs, Diagonalization Limits, Space Complexity January 22, 2016 Lecturer: Paul Beame Scribe: Paul Beame Diagonalization enabled us to separate

More information

Lecture 3: Reductions and Completeness

Lecture 3: Reductions and Completeness CS 710: Complexity Theory 9/13/2011 Lecture 3: Reductions and Completeness Instructor: Dieter van Melkebeek Scribe: Brian Nixon Last lecture we introduced the notion of a universal Turing machine for deterministic

More information

The Size and Depth of Layered Boolean Circuits

The Size and Depth of Layered Boolean Circuits The Size and Depth of Layered Boolean Circuits Anna Gál a,1, Jing-Tang Jang a,2 a Dept. of Computer Science, University of Texas at Austin, Austin, TX 78712-1188, USA Abstract We consider the relationship

More information

NP, polynomial-time mapping reductions, and NP-completeness

NP, polynomial-time mapping reductions, and NP-completeness NP, polynomial-time mapping reductions, and NP-completeness In the previous lecture we discussed deterministic time complexity, along with the time-hierarchy theorem, and introduced two complexity classes:

More information

There Is No Polynomial Deterministic Space Simulation of Probabilistic Space with a Two-Way Random-Tape Generator

There Is No Polynomial Deterministic Space Simulation of Probabilistic Space with a Two-Way Random-Tape Generator INFORMATION AND CONTROL 67, 158-162 (1985) There Is No Polynomial Deterministic Space Simulation of Probabilistic Space with a Two-Way Random-Tape Generator MAREK KARPINSKI* Department of Computer Science,

More information

Space and Nondeterminism

Space and Nondeterminism CS 221 Computational Complexity, Lecture 5 Feb 6, 2018 Space and Nondeterminism Instructor: Madhu Sudan 1 Scribe: Yong Wook Kwon Topic Overview Today we ll talk about space and non-determinism. For some

More information

Time-Space Tradeoffs for SAT

Time-Space Tradeoffs for SAT Lecture 8 Time-Space Tradeoffs for SAT April 22, 2004 Lecturer: Paul Beame Notes: Definition 8.1. TIMESPACE(T (n), S(n)) = {L {0, 1} offline, multitape TM M such that L = L(M) and M uses time O(T (n))

More information

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

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

More information

1 Computational Problems

1 Computational Problems Stanford University CS254: Computational Complexity Handout 2 Luca Trevisan March 31, 2010 Last revised 4/29/2010 In this lecture we define NP, we state the P versus NP problem, we prove that its formulation

More information

1 PSPACE-Completeness

1 PSPACE-Completeness CS 6743 Lecture 14 1 Fall 2007 1 PSPACE-Completeness Recall the NP-complete problem SAT: Is a given Boolean formula φ(x 1,..., x n ) satisfiable? The same question can be stated equivalently as: Is the

More information

A Short History of Computational Complexity

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

More information

Kernelization by matroids: Odd Cycle Transversal

Kernelization by matroids: Odd Cycle Transversal Lecture 8 (10.05.2013) Scribe: Tomasz Kociumaka Lecturer: Marek Cygan Kernelization by matroids: Odd Cycle Transversal 1 Introduction The main aim of this lecture is to give a polynomial kernel for the

More information

Length-Increasing Reductions for PSPACE-Completeness

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

More information

Notes for Lecture Notes 2

Notes for Lecture Notes 2 Stanford University CS254: Computational Complexity Notes 2 Luca Trevisan January 11, 2012 Notes for Lecture Notes 2 In this lecture we define NP, we state the P versus NP problem, we prove that its formulation

More information

Boolean complexity classes vs. their arithmetic analogs

Boolean complexity classes vs. their arithmetic analogs Boolean complexity classes vs. their arithmetic analogs Anna Gál Avi Wigderson Abstract This paper provides logspace and small circuit depth analogs of the result of Valiant and Vazirani, which is a randomized

More information

Lecture 59 : Instance Compression and Succinct PCP s for NP

Lecture 59 : Instance Compression and Succinct PCP s for NP IITM-CS6840: Advanced Complexity Theory March 31, 2012 Lecture 59 : Instance Compression and Succinct PCP s for NP Lecturer: Sivaramakrishnan N.R. Scribe: Prashant Vasudevan 1 Introduction Classical Complexity

More information

198:538 Complexity of Computation Lecture 16 Rutgers University, Spring March 2007

198:538 Complexity of Computation Lecture 16 Rutgers University, Spring March 2007 198:538 Complexity of Computation Lecture 16 Rutgers University, Spring 2007 8 March 2007 In this lecture we discuss Shamir s theorem that PSPACE is the set of languages that have interactive proofs with

More information

Generalized Lowness and Highness and Probabilistic Complexity Classes

Generalized Lowness and Highness and Probabilistic Complexity Classes Generalized Lowness and Highness and Probabilistic Complexity Classes Andrew Klapper University of Manitoba Abstract We introduce generalized notions of low and high complexity classes and study their

More information

SOLUTION: SOLUTION: SOLUTION:

SOLUTION: SOLUTION: SOLUTION: Convert R and S into nondeterministic finite automata N1 and N2. Given a string s, if we know the states N1 and N2 may reach when s[1...i] has been read, we are able to derive the states N1 and N2 may

More information

Logarithmic space. Evgenij Thorstensen V18. Evgenij Thorstensen Logarithmic space V18 1 / 18

Logarithmic space. Evgenij Thorstensen V18. Evgenij Thorstensen Logarithmic space V18 1 / 18 Logarithmic space Evgenij Thorstensen V18 Evgenij Thorstensen Logarithmic space V18 1 / 18 Journey below Unlike for time, it makes sense to talk about sublinear space. This models computations on input

More information

Data Structures in Java

Data Structures in Java Data Structures in Java Lecture 21: Introduction to NP-Completeness 12/9/2015 Daniel Bauer Algorithms and Problem Solving Purpose of algorithms: find solutions to problems. Data Structures provide ways

More information

Universality for Nondeterministic Logspace

Universality for Nondeterministic Logspace Universality for Nondeterministic Logspace Vinay Chaudhary, Anand Kumar Sinha, and Somenath Biswas Department of Computer Science and Engineering IIT Kanpur September 2004 1 Introduction The notion of

More information

Lecture Space Bounded and Random Turing Machines and RL

Lecture Space Bounded and Random Turing Machines and RL 6.84 Randomness and Computation October 11, 017 Lecture 10 Lecturer: Ronitt Rubinfeld Scribe: Tom Kolokotrones 1 Space Bounded and Random Turing Machines and RL 1.1 Space Bounded and Random Turing Machines

More information

Brady Garvin, Derrick Stolee, Raghunath Tewari, and N. V. Vinodchandran

Brady Garvin, Derrick Stolee, Raghunath Tewari, and N. V. Vinodchandran comput. complex. 23 (2014), 85 98 c Springer Basel 2012 1016-3328/14/010085-14 published online November 6, 2012 DOI 10.1007/s00037-012-0050-8 computational complexity REACHFEWL =REACHUL Brady Garvin,

More information

ON TRAVERSAL SEQUENCES, EXPLORATION SEQUENCES AND COMPLETENESS OF KOLMOGOROV RANDOM STRINGS

ON TRAVERSAL SEQUENCES, EXPLORATION SEQUENCES AND COMPLETENESS OF KOLMOGOROV RANDOM STRINGS ON TRAVERSAL SEQUENCES, EXPLORATION SEQUENCES AND COMPLETENESS OF KOLMOGOROV RANDOM STRINGS BY MICHAL KOUCKÝ A dissertation submitted to the Graduate School New Brunswick Rutgers, The State University

More information

Parallelism and Machine Models

Parallelism and Machine Models Parallelism and Machine Models Andrew D Smith University of New Brunswick, Fredericton Faculty of Computer Science Overview Part 1: The Parallel Computation Thesis Part 2: Parallelism of Arithmetic RAMs

More information

Randomness and non-uniformity

Randomness and non-uniformity Randomness and non-uniformity JASS 2006 Course 1: Proofs and Computers Felix Weninger TU München April 2006 Outline Randomized computation 1 Randomized computation 2 Computation with advice Non-uniform

More information

Pseudorandom Generators for Regular Branching Programs

Pseudorandom Generators for Regular Branching Programs Pseudorandom Generators for Regular Branching Programs Mark Braverman Anup Rao Ran Raz Amir Yehudayoff Abstract We give new pseudorandom generators for regular read-once branching programs of small width.

More information

1 Computational problems

1 Computational problems 80240233: Computational Complexity Lecture 1 ITCS, Tsinghua Univesity, Fall 2007 9 October 2007 Instructor: Andrej Bogdanov Notes by: Andrej Bogdanov The aim of computational complexity theory is to study

More information

COMPUTATIONAL COMPLEXITY

COMPUTATIONAL COMPLEXITY COMPUTATIONAL COMPLEXITY A Modern Approach SANJEEV ARORA Princeton University BOAZ BARAK Princeton University {Щ CAMBRIDGE Щ0 UNIVERSITY PRESS Contents About this book Acknowledgments Introduction page

More information

6.841/18.405J: Advanced Complexity Wednesday, February 12, Lecture Lecture 3

6.841/18.405J: Advanced Complexity Wednesday, February 12, Lecture Lecture 3 6.841/18.405J: Advanced Complexity Wednesday, February 12, 2003 Lecture Lecture 3 Instructor: Madhu Sudan Scribe: Bobby Kleinberg 1 The language MinDNF At the end of the last lecture, we introduced the

More information

The diameter of randomly perturbed digraphs and some applications

The diameter of randomly perturbed digraphs and some applications The diameter of randomly perturbed digraphs and some applications Abraham D. Flaxman and Alan M. Frieze Department of Mathematical Sciences, Carnegie Mellon University, Pittsburgh, PA, 15213, USA abie@cmu.edu,

More information

Balanced Allocation Through Random Walk

Balanced Allocation Through Random Walk Balanced Allocation Through Random Walk Alan Frieze Samantha Petti November 25, 2017 Abstract We consider the allocation problem in which m (1 ε)dn items are to be allocated to n bins with capacity d.

More information

Lecture 5: The Landscape of Complexity Classes

Lecture 5: The Landscape of Complexity Classes IAS/PCMI Summer Session 2000 Clay Mathematics Undergraduate Program Basic Course on Computational Complexity Lecture 5: The Landscape of Complexity Classes David Mix Barrington and Alexis Maciel July 21,

More information

Quantum Algorithms for Evaluating Min-Max Trees

Quantum Algorithms for Evaluating Min-Max Trees Quantum Algorithms for Evaluating Min-Max Trees Richard Cleve 1,2,DmitryGavinsky 1, and D. L. Yonge-Mallo 1 1 David R. Cheriton School of Computer Science and Institute for Quantum Computing, University

More information

Probabilistically Checkable Proofs. 1 Introduction to Probabilistically Checkable Proofs

Probabilistically Checkable Proofs. 1 Introduction to Probabilistically Checkable Proofs Course Proofs and Computers, JASS 06 Probabilistically Checkable Proofs Lukas Bulwahn May 21, 2006 1 Introduction to Probabilistically Checkable Proofs 1.1 History of Inapproximability Results Before introducing

More information

1 Deterministic Turing Machines

1 Deterministic Turing Machines Time and Space Classes Exposition by William Gasarch 1 Deterministic Turing Machines Turing machines are a model of computation. It is believed that anything that can be computed can be computed by a Turing

More information

fsat We next show that if sat P, then fsat has a polynomial-time algorithm. c 2010 Prof. Yuh-Dauh Lyuu, National Taiwan University Page 425

fsat We next show that if sat P, then fsat has a polynomial-time algorithm. c 2010 Prof. Yuh-Dauh Lyuu, National Taiwan University Page 425 fsat fsat is this function problem: Let φ(x 1, x 2,..., x n ) be a boolean expression. If φ is satisfiable, then return a satisfying truth assignment. Otherwise, return no. We next show that if sat P,

More information

2-LOCAL RANDOM REDUCTIONS TO 3-VALUED FUNCTIONS

2-LOCAL RANDOM REDUCTIONS TO 3-VALUED FUNCTIONS 2-LOCAL RANDOM REDUCTIONS TO 3-VALUED FUNCTIONS A. Pavan and N. V. Vinodchandran Abstract. Yao (in a lecture at DIMACS Workshop on structural complexity and cryptography, 1990) showed that if a language

More information

Space Complexity. The space complexity of a program is how much memory it uses.

Space Complexity. The space complexity of a program is how much memory it uses. Space Complexity The space complexity of a program is how much memory it uses. Measuring Space When we compute the space used by a TM, we do not count the input (think of input as readonly). We say that

More information

Notes for Lecture 3... x 4

Notes for Lecture 3... x 4 Stanford University CS254: Computational Complexity Notes 3 Luca Trevisan January 18, 2012 Notes for Lecture 3 In this lecture we introduce the computational model of boolean circuits and prove that polynomial

More information

COMPUTATIONAL COMPLEXITY

COMPUTATIONAL COMPLEXITY ATHEATICS: CONCEPTS, AND FOUNDATIONS Vol. III - Computational Complexity - Osamu Watanabe COPUTATIONAL COPLEXITY Osamu Watanabe Tokyo Institute of Technology, Tokyo, Japan Keywords: {deterministic, randomized,

More information

The power and weakness of randomness (when you are short on time) Avi Wigderson Institute for Advanced Study

The power and weakness of randomness (when you are short on time) Avi Wigderson Institute for Advanced Study The power and weakness of randomness (when you are short on time) Avi Wigderson Institute for Advanced Study Plan of the talk Computational complexity -- efficient algorithms, hard and easy problems, P

More information

Computer Sciences Department

Computer Sciences Department Computer Sciences Department 1 Reference Book: INTRODUCTION TO THE THEORY OF COMPUTATION, SECOND EDITION, by: MICHAEL SIPSER Computer Sciences Department 3 ADVANCED TOPICS IN C O M P U T A B I L I T Y

More information

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

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

More information

Notes on Complexity Theory Last updated: October, Lecture 6

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

More information

Pseudorandom Generators

Pseudorandom Generators 8 Pseudorandom Generators Great Ideas in Theoretical Computer Science Saarland University, Summer 2014 andomness is one of the fundamental computational resources and appears everywhere. In computer science,

More information

INSTITUTE OF MATHEMATICS THE CZECH ACADEMY OF SCIENCES. A trade-off between length and width in resolution. Neil Thapen

INSTITUTE OF MATHEMATICS THE CZECH ACADEMY OF SCIENCES. A trade-off between length and width in resolution. Neil Thapen INSTITUTE OF MATHEMATICS THE CZECH ACADEMY OF SCIENCES A trade-off between length and width in resolution Neil Thapen Preprint No. 59-2015 PRAHA 2015 A trade-off between length and width in resolution

More information

Deterministic approximation for the cover time of trees

Deterministic approximation for the cover time of trees Deterministic approximation for the cover time of trees Uriel Feige Ofer Zeitouni September 10, 2009 Abstract We present a deterministic algorithm that given a tree T with n vertices, a starting vertex

More information

Lecture 8: Complete Problems for Other Complexity Classes

Lecture 8: Complete Problems for Other Complexity Classes IAS/PCMI Summer Session 2000 Clay Mathematics Undergraduate Program Basic Course on Computational Complexity Lecture 8: Complete Problems for Other Complexity Classes David Mix Barrington and Alexis Maciel

More information

There are no zero-hard problems in multiparty communication complexity

There are no zero-hard problems in multiparty communication complexity There are no zero-hard problems in multiparty communication complexity László Babai and Denis Pankratov University of Chicago DRAFT: 2015-04-29 7:00pm Abstract We study the feasibility of generalizing

More information

Notes on Complexity Theory Last updated: November, Lecture 10

Notes on Complexity Theory Last updated: November, Lecture 10 Notes on Complexity Theory Last updated: November, 2015 Lecture 10 Notes by Jonathan Katz, lightly edited by Dov Gordon. 1 Randomized Time Complexity 1.1 How Large is BPP? We know that P ZPP = RP corp

More information

Lower Bounds for Dynamic Connectivity (2004; Pǎtraşcu, Demaine)

Lower Bounds for Dynamic Connectivity (2004; Pǎtraşcu, Demaine) Lower Bounds for Dynamic Connectivity (2004; Pǎtraşcu, Demaine) Mihai Pǎtraşcu, MIT, web.mit.edu/ mip/www/ Index terms: partial-sums problem, prefix sums, dynamic lower bounds Synonyms: dynamic trees 1

More information

Lecture 10 Algorithmic version of the local lemma

Lecture 10 Algorithmic version of the local lemma Lecture 10 Algorithmic version of the local lemma Uriel Feige Department of Computer Science and Applied Mathematics The Weizman Institute Rehovot 76100, Israel uriel.feige@weizmann.ac.il June 9, 2014

More information

Separating the Power of EREW and CREW PRAMs with Small Communication Width*

Separating the Power of EREW and CREW PRAMs with Small Communication Width* information and computation 138, 8999 (1997) article no. IC97649 Separating the Power of EREW and CREW PRAMs with Small Communication Width* Paul Beame Department of Computer Science and Engineering, University

More information

1 Review of Vertex Cover

1 Review of Vertex Cover CS266: Parameterized Algorithms and Complexity Stanford University Lecture 3 Tuesday, April 9 Scribe: Huacheng Yu Spring 2013 1 Review of Vertex Cover In the last lecture, we discussed FPT algorithms for

More information

Computational Complexity Theory. Markus Bläser, Holger Dell, Karteek Sreenivasaiah Universität des Saarlandes Draft June 15, 2015 and forever

Computational Complexity Theory. Markus Bläser, Holger Dell, Karteek Sreenivasaiah Universität des Saarlandes Draft June 15, 2015 and forever Computational Complexity Theory Markus Bläser, Holger Dell, Karteek Sreenivasaiah Universität des Saarlandes Draft June 15, 2015 and forever 2 1 Simple lower bounds and gaps Complexity theory is the science

More information

Lecture 24: Randomized Complexity, Course Summary

Lecture 24: Randomized Complexity, Course Summary 6.045 Lecture 24: Randomized Complexity, Course Summary 1 1/4 1/16 1/4 1/4 1/32 1/16 1/32 Probabilistic TMs 1/16 A probabilistic TM M is a nondeterministic TM where: Each nondeterministic step is called

More information

Polynomial Hierarchy

Polynomial Hierarchy Polynomial Hierarchy A polynomial-bounded version of Kleene s Arithmetic Hierarchy becomes trivial if P = NP. Karp, 1972 Computational Complexity, by Fu Yuxi Polynomial Hierarchy 1 / 44 Larry Stockmeyer

More information

arxiv: v2 [cs.ds] 3 Oct 2017

arxiv: v2 [cs.ds] 3 Oct 2017 Orthogonal Vectors Indexing Isaac Goldstein 1, Moshe Lewenstein 1, and Ely Porat 1 1 Bar-Ilan University, Ramat Gan, Israel {goldshi,moshe,porately}@cs.biu.ac.il arxiv:1710.00586v2 [cs.ds] 3 Oct 2017 Abstract

More information

hal , version 1-3 Feb 2008

hal , version 1-3 Feb 2008 Author manuscript, published in "STACS 2008, Bordeaux : France (2008)" Symposium on Theoretical Aspects of Computer Science 2008 (Bordeaux), pp. 433-444 www.stacs-conf.org SPACE HIERARCHY RESULTS FOR RANDOMIZED

More information

1 Introduction (January 21)

1 Introduction (January 21) CS 97: Concrete Models of Computation Spring Introduction (January ). Deterministic Complexity Consider a monotonically nondecreasing function f : {,,..., n} {, }, where f() = and f(n) =. We call f a step

More information

Universal Algebra and Computational Complexity Lecture 1

Universal Algebra and Computational Complexity Lecture 1 Universal Algebra and Computational Complexity Lecture 1 Ross Willard University of Waterloo, Canada Třešt, September 2008 Ross Willard (Waterloo) Algebra and Complexity Třešt, September 2008 1 / 23 Outline

More information

Chicago Journal of Theoretical Computer Science. MIT Press

Chicago Journal of Theoretical Computer Science. MIT Press Chicago Journal of Theoretical Computer Science MIT Press Volume 1995, Article 1 30 June, 1995 ISSN 1073 0486. MIT Press Journals, 55 Hayward St., Cambridge, MA 02142; (617)253-2889; journals-orders@mit.edu,

More information

Planning With Information States: A Survey Term Project for cs397sml Spring 2002

Planning With Information States: A Survey Term Project for cs397sml Spring 2002 Planning With Information States: A Survey Term Project for cs397sml Spring 2002 Jason O Kane jokane@uiuc.edu April 18, 2003 1 Introduction Classical planning generally depends on the assumption that the

More information