-
Online Matching with High Probability
Authors:
Milena Mihail,
Thorben Tröbst
Abstract:
We study the classical, randomized Ranking algorithm which is known to be $(1 - \frac{1}{e})$-competitive in expectation for the Online Bipartite Matching Problem. We give a tail inequality bound, namely that Ranking is $(1 - \frac{1}{e} - α)$-competitive with probability at least $1 - e^{-2 α^2 n}$ where $n$ is the size of the maximum matching in the instance. Building on this, we show similar co…
▽ More
We study the classical, randomized Ranking algorithm which is known to be $(1 - \frac{1}{e})$-competitive in expectation for the Online Bipartite Matching Problem. We give a tail inequality bound, namely that Ranking is $(1 - \frac{1}{e} - α)$-competitive with probability at least $1 - e^{-2 α^2 n}$ where $n$ is the size of the maximum matching in the instance. Building on this, we show similar concentration results for the Fully Online Matching Problem and for the Online Vertex-Weighted Bipartite Matching Problem.
△ Less
Submitted 14 December, 2021;
originally announced December 2021.
-
Graphic Realizations of Joint-Degree Matrices
Authors:
Georgios Amanatidis,
Bradley Green,
Milena Mihail
Abstract:
In this paper we introduce extensions and modifications of the classical degree sequence graphic realization problem studied by Erdős-Gallai and Havel-Hakimi, as well as of the corresponding connected graphic realization version. We define the joint-degree matrix graphic (resp. connected graphic) realization problem, where in addition to the degree sequence, the exact number of desired edges betwe…
▽ More
In this paper we introduce extensions and modifications of the classical degree sequence graphic realization problem studied by Erdős-Gallai and Havel-Hakimi, as well as of the corresponding connected graphic realization version. We define the joint-degree matrix graphic (resp. connected graphic) realization problem, where in addition to the degree sequence, the exact number of desired edges between vertices of different degree classes is also specified. We give necessary and sufficient conditions, and polynomial time decision and construction algorithms for the graphic and connected graphic realization problems. These problems arise naturally in the current topic of graph modeling for complex networks. From the technical point of view, the joint-degree matrix realization algorithm is straightforward. However, the connected joint-degree matrix realization algorithm involves a novel recursive search of suitable local graph modifications. Also, we outline directions for further work of both theoretical and practical interest. In particular, we give a Markov chain which converges to the uniform distribution over all realizations. We show that the underline state space is connected, and we leave the question of the mixing rate open.
△ Less
Submitted 23 September, 2015;
originally announced September 2015.
-
Efficient Generation ε-close to G(n,p) and Generalizations
Authors:
Antonio Blanca,
Milena Mihail
Abstract:
We give an efficient algorithm to generate a graph from a distribution $ε$-close to $G(n,p)$, in the sense of total variation distance. In particular, if $p$ is represented with $O(\log n)$-bit accuracy, then, with high probability, the running time is linear in the expected number of edges of the output graph (up to poly-logarithmic factors). All our running times include the complexity of the ar…
▽ More
We give an efficient algorithm to generate a graph from a distribution $ε$-close to $G(n,p)$, in the sense of total variation distance. In particular, if $p$ is represented with $O(\log n)$-bit accuracy, then, with high probability, the running time is linear in the expected number of edges of the output graph (up to poly-logarithmic factors). All our running times include the complexity of the arithmetic involved in the corresponding algorithms. Previous standard methods for exact $G(n,p)$ sampling (see e.g. Batagelj and Brandes, 2005) achieve similar running times, however, under the assumption that performing real number arithmetic with arbitrary accuracy takes constant time. We note that the actual accuracy required by these methods is O(n)-bit per step, which results in quadratic running times.
The main idea of our $G(n,p)$ generation algorithm is a Metropolis Markov chain to sample $ε$-close from the binomial distribution. This is a new method for sampling from the binomial distribution: it is of separate interest and may find other useful applications. Our analysis accounts for all necessary bit-accuracy and arithmetic, and our running times are comparable to known methods for exact binomial sampling.
We further obtain efficient generation algorithms for random graphs with given arbitrary degree distributions, Inhomogeneous Random Graphs when the kernel function is the inner product, and Stochastic Kronecker Graphs. To the best our knowledge, our work can be viewed as the first effort to simulate efficient generation of graphs from classical random graph models, while taking into account implementational considerations as fundamental computational aspects, and quantifying the tradeoff between accuracy and running time in a way that can be useful in practice.
△ Less
Submitted 12 July, 2012; v1 submitted 26 April, 2012;
originally announced April 2012.
