Improved lower bounds on the randomized complexity of graph properties

Amit Chakrabarti, Subhash Khot

Research output: Contribution to journalArticle

Abstract

We prove a lower bound of Ω(n4/3 log1/3 n) on the randomized decision tree complexity of any nontrivial monotone n-vertex graph property, and of any nontrivial monotone bipartite graph property with bipartitions of size n. This improves the previous best bound of Ω(n 4/3) due to Hajnal (Combinatorica 11 (1991) 131-143). Our proof works by improving a graph packing lemma used in earlier work, and this improvement in turn stems from a novel probabilistic analysis. Graph packing being a well-studied subject in its own right, our improved packing lemma and the probabilistic technique used to prove it may be of independent interest.

Original languageEnglish (US)
Pages (from-to)427-440
Number of pages14
JournalRandom Structures and Algorithms
Volume30
Issue number3
DOIs
StatePublished - May 2007

Fingerprint

Graph Packing
Lemma
Monotone
Lower bound
Probabilistic Analysis
Graph in graph theory
Decision tree
Bipartite Graph
Packing
Decision trees
Vertex of a graph

Keywords

  • Complexity
  • Decision trees
  • Graph packing
  • Graph properties
  • Probabilistic method
  • Randomized algorithms

ASJC Scopus subject areas

  • Computer Graphics and Computer-Aided Design
  • Software
  • Mathematics(all)
  • Applied Mathematics

Cite this

Improved lower bounds on the randomized complexity of graph properties. / Chakrabarti, Amit; Khot, Subhash.

In: Random Structures and Algorithms, Vol. 30, No. 3, 05.2007, p. 427-440.

Research output: Contribution to journalArticle

@article{343517a0558e4c56820dc2f9ec237ca0,
title = "Improved lower bounds on the randomized complexity of graph properties",
abstract = "We prove a lower bound of Ω(n4/3 log1/3 n) on the randomized decision tree complexity of any nontrivial monotone n-vertex graph property, and of any nontrivial monotone bipartite graph property with bipartitions of size n. This improves the previous best bound of Ω(n 4/3) due to Hajnal (Combinatorica 11 (1991) 131-143). Our proof works by improving a graph packing lemma used in earlier work, and this improvement in turn stems from a novel probabilistic analysis. Graph packing being a well-studied subject in its own right, our improved packing lemma and the probabilistic technique used to prove it may be of independent interest.",
keywords = "Complexity, Decision trees, Graph packing, Graph properties, Probabilistic method, Randomized algorithms",
author = "Amit Chakrabarti and Subhash Khot",
year = "2007",
month = "5",
doi = "10.1002/rsa.20164",
language = "English (US)",
volume = "30",
pages = "427--440",
journal = "Random Structures and Algorithms",
issn = "1042-9832",
publisher = "John Wiley and Sons Ltd",
number = "3",

}

TY - JOUR

T1 - Improved lower bounds on the randomized complexity of graph properties

AU - Chakrabarti, Amit

AU - Khot, Subhash

PY - 2007/5

Y1 - 2007/5

N2 - We prove a lower bound of Ω(n4/3 log1/3 n) on the randomized decision tree complexity of any nontrivial monotone n-vertex graph property, and of any nontrivial monotone bipartite graph property with bipartitions of size n. This improves the previous best bound of Ω(n 4/3) due to Hajnal (Combinatorica 11 (1991) 131-143). Our proof works by improving a graph packing lemma used in earlier work, and this improvement in turn stems from a novel probabilistic analysis. Graph packing being a well-studied subject in its own right, our improved packing lemma and the probabilistic technique used to prove it may be of independent interest.

AB - We prove a lower bound of Ω(n4/3 log1/3 n) on the randomized decision tree complexity of any nontrivial monotone n-vertex graph property, and of any nontrivial monotone bipartite graph property with bipartitions of size n. This improves the previous best bound of Ω(n 4/3) due to Hajnal (Combinatorica 11 (1991) 131-143). Our proof works by improving a graph packing lemma used in earlier work, and this improvement in turn stems from a novel probabilistic analysis. Graph packing being a well-studied subject in its own right, our improved packing lemma and the probabilistic technique used to prove it may be of independent interest.

KW - Complexity

KW - Decision trees

KW - Graph packing

KW - Graph properties

KW - Probabilistic method

KW - Randomized algorithms

UR - http://www.scopus.com/inward/record.url?scp=34247525606&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34247525606&partnerID=8YFLogxK

U2 - 10.1002/rsa.20164

DO - 10.1002/rsa.20164

M3 - Article

VL - 30

SP - 427

EP - 440

JO - Random Structures and Algorithms

JF - Random Structures and Algorithms

SN - 1042-9832

IS - 3

ER -