Complexity of finding dense subgraphs

被引:79
作者
Asahiro, Y [1 ]
Hassin, R
Iwama, K
机构
[1] Kyushu Univ, Dept Comp Sci & Commun Engn, Fukuoka 8128581, Japan
[2] Tel Aviv Univ, Dept Stat & Operat Res, IL-69978 Tel Aviv, Israel
[3] Kyoto Univ, Sch Informat, Kyoto 6068501, Japan
关键词
dense subgraph; regular graph; NP-complete; threshold of complexity;
D O I
10.1016/S0166-218X(01)00243-8
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
The k-f (k) dense subgraph problem((k, f (k))-DSP) asks whether there is a k-vertex subgraph of a given graph G which has at least f (k) edges. When f (k) = k(k - 1)/2, (k, f (k))-DSP is equivalent to the well-known k-clique problem. The main purpose of this paper is to discuss the problem of finding slightly dense subgraphs. Note that f(k) is about k(2) for the k-clique problem. It is shown that (k, f (k))-DSP remains NP-complete for f (k) = Theta(k(1-epsilon)) where epsilon may be any constant such that 0 < epsilon < 1. It is also NP-complete for "relatively" slightly-dense subgraphs, i.e., (k, f(k))-DSP is NP-complete for f(k) = ek(2)/v(2)(1 +O(v(epsilon-1))), where v is the number of G's vertices and e is the number of G's edges. This condition is quite tight because the answer to (k, f (k))-DSP is always yes for f (k) = ek(2/)v(2) (1 - (v - k)/(vk - k)) that is the average number of edges in a subgraph of k vertices. Also, we show that the hardness of (k, f(k))-DSP remains for regular graphs: (k, f (k))-DSP is NP-complete for Theta(v(epsilon))-regular graphs if f (k) = Theta(k(1+epsilon2)) for any 0 < epsilon(1), epsilon(2) < 1. (C) 2002 Elsevier Science B.V. All rights reserved.
引用
收藏
页码:15 / 26
页数:12
相关论文
共 14 条
  • [1] ALON N, 1992, PROBABILISTIC METHOD
  • [2] [Anonymous], 1996, LNCS
  • [3] Arora S., 1995, Proceedings of the Twenty-Seventh Annual ACM Symposium on the Theory of Computing, P284, DOI 10.1145/225058.225140
  • [4] Cook S.A., 1971, P 3 ANN ACM S THEOR, P151, DOI DOI 10.1145/800157.805047
  • [5] Maximum dispersion problem in dense graphs
    Czygrinow, A
    [J]. OPERATIONS RESEARCH LETTERS, 2000, 27 (05) : 223 - 227
  • [6] FEIGE U, CS9716 FAC MATH COMP
  • [7] FEIGE U, UNPUB DENSE K SUBGRA
  • [8] A FAST PARAMETRIC MAXIMUM FLOW ALGORITHM AND APPLICATIONS
    GALLO, G
    GRIGORIADIS, MD
    TARJAN, RE
    [J]. SIAM JOURNAL ON COMPUTING, 1989, 18 (01) : 30 - 55
  • [9] Garey M. R., 1976, Theoretical Computer Science, V1, P237, DOI 10.1016/0304-3975(76)90059-1
  • [10] Approximation algorithms for maximum dispersion
    Hassin, R
    Rubinstein, S
    Tamir, A
    [J]. OPERATIONS RESEARCH LETTERS, 1997, 21 (03) : 133 - 137