2004
DOI: 10.1002/rsa.20008
|View full text |Cite
|
Sign up to set email alerts
|

Large deviations for sums of partly dependent random variables

Abstract: Abstract. We use and extend a method by Hoeffding to obtain strong large deviation bounds for sums of dependent random variables with suitable dependency structure. The method is based on breaking up the sum into sums of independent variables. Applications are given to U -statistics, random strings and random graphs.

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

2
199
0
1

Year Published

2009
2009
2016
2016

Publication Types

Select...
5
2
1

Relationship

0
8

Authors

Journals

citations
Cited by 147 publications
(203 citation statements)
references
References 20 publications
(32 reference statements)
2
199
0
1
Order By: Relevance
“…In conclusion, the combination of (6) and (8) implies that, for large n, algorithm A outputs a nearly balanced recut on Π (G, * ) with high probability. By the discussion in Sect.…”
Section: Local Concentration Boundmentioning
confidence: 86%
See 1 more Smart Citation
“…In conclusion, the combination of (6) and (8) implies that, for large n, algorithm A outputs a nearly balanced recut on Π (G, * ) with high probability. By the discussion in Sect.…”
Section: Local Concentration Boundmentioning
confidence: 86%
“…Since the maximum degree of G 2r is at most max v |B G (v, 2r)| = o(n), this graph can always be partitioned into χ(G 2r ) = o(n) independent sets. Indeed, Janson [8] presents large deviation bounds for sums of type (7) by applying Chernoff-Hoeffding bounds for each colour class in a χ(G 2r )-colouring of G 2r . For any ε > 0, Theorem 2.1 in Janson [8], as applied to our setting, gives…”
Section: Local Concentration Boundmentioning
confidence: 99%
“…Applications of the Hajnal-Szemerédi theorem and recent results on equitable colorings of graphs can be found in (among others) [1], [2], [9], [11], [12], [19]. Equitable coloring turned out to be useful in establishing bounds on tails of sums of dependent variables [6], [8], [18].…”
Section: Introduction and Main Resultsmentioning
confidence: 99%
“…Let X follow a binomial distribution with parameters n and p. Then and P(X np + t) exp − t Proofs of these estimates may be found in [25]; see also [3] and [1]. They can be derived from an application of Markov's inequality to the random variable e λ(X−np) using the fact that the moment generating function Ee λ(X−np) is e −λpn (pe λ + 1 − p) n .…”
Section: 2mentioning
confidence: 99%