اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدTowards Tight Bounds for Spectral Sparsification of Hypergraphs
93
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Cut and spectral sparsification of graphs have numerous applications, including e.g. speeding up algorithms for cuts and Laplacian solvers. These powerful notions have recently been extended to hypergraphs, which are much richer and may offer new applications. However, the current bounds on the size of hypergraph sparsifiers are not as tight as the corresponding bounds for graphs. Our first result is a polynomial-time algorithm that, given a hypergraph on $n$ vertices with maximum hyperedge size $r$, outputs an $epsilon$-spectral sparsifier with $O^*(nr)$ hyperedges, where $O^*$ suppresses $(epsilon^{-1} log n)^{O(1)}$ factors. This size bound improves the two previous bounds: $O^*(n^3)$ [Soma and Yoshida, SODA19] and $O^*(nr^3)$ [Bansal, Svensson and Trevisan, FOCS19]. Our main technical tool is a new method for proving concentration of the nonlinear analogue of the quadratic form of the Laplacians for hypergraph expanders. We complement this with lower bounds on the bit complexity of any compression scheme that $(1+epsilon)$-approximates all the cuts in a given hypergraph, and hence also on the bit complexity of every $epsilon$-cut/spectral sparsifier. These lower bounds are based on Ruzsa-Szemeredi graphs, and a particular instantiation yields an $Omega(nr)$ lower bound on the bit complexity even for fixed constant $epsilon$. This is tight up to polylogarithmic factors in $n$, due to recent hypergraph cut sparsifiers of [Chen, Khanna and Nagda, FOCS20]. Finally, for directed hypergraphs, we present an algorithm that computes an $epsilon$-spectral sparsifier with $O^*(n^2r^3)$ hyperarcs, where $r$ is the maximum size of a hyperarc. For small $r$, this improves over $O^*(n^3)$ known from [Soma and Yoshida, SODA19], and is getting close to the trivial lower bound of $Omega(n^2)$ hyperarcs.
قيم البحث
اقرأ أيضاً
Among the most important graph parameters is the Diameter, the largest distance between any two vertices. There are no known very efficient algorithms for computing the Diameter exactly. Thus, much research has been devoted to how fast this parameter
can be approximated. Chechik et al. showed that the diameter can be approximated within a multiplicative factor of $3/2$ in $tilde{O}(m^{3/2})$ time. Furthermore, Roditty and Vassilevska W. showed that unless the Strong Exponential Time Hypothesis (SETH) fails, no $O(n^{2-epsilon})$ time algorithm can achieve an approximation factor better than $3/2$ in sparse graphs. Thus the above algorithm is essentially optimal for sparse graphs for approximation factors less than $3/2$. It was, however, completely plausible that a $3/2$-approximation is possible in linear time. In this work we conditionally rule out such a possibility by showing that unless SETH fails no $O(m^{3/2-epsilon})$ time algorithm can achieve an approximation factor better than $5/3$. Another fundamental set of graph parameters are the Eccentricities. The Eccentricity of a vertex $v$ is the distance between $v$ and the farthest vertex from $v$. Chechik et al. showed that the Eccentricities of all vertices can be approximated within a factor of $5/3$ in $tilde{O}(m^{3/2})$ time and Abboud et al. showed that no $O(n^{2-epsilon})$ algorithm can achieve better than $5/3$ approximation in sparse graphs. We show that the runtime of the $5/3$ approximation algorithm is also optimal under SETH. We also show that no near-linear time algorithm can achieve a better than $2$ approximation for the Eccentricities and that this is essentially tight: we give an algorithm that approximates Eccentricities within a $2+delta$ factor in $tilde{O}(m/delta)$ time for any $0<delta<1$. This beats all Eccentricity algorithms in Cairo et al.
Vizings celebrated theorem asserts that any graph of maximum degree $Delta$ admits an edge coloring using at most $Delta+1$ colors. In contrast, Bar-Noy, Naor and Motwani showed over a quarter century that the trivial greedy algorithm, which uses $2D
elta-1$ colors, is optimal among online algorithms. Their lower bound has a caveat, however: it only applies to low-degree graphs, with $Delta=O(log n)$, and they conjectured the existence of online algorithms using $Delta(1+o(1))$ colors for $Delta=omega(log n)$. Progress towards resolving this conjecture was only made under stochastic arrivals (Aggarwal et al., FOCS03 and Bahmani et al., SODA10). We resolve the above conjecture for emph{adversarial} vertex arrivals in bipartite graphs, for which we present a $(1+o(1))Delta$-edge-coloring algorithm for $Delta=omega(log n)$ known a priori. Surprisingly, if $Delta$ is not known ahead of time, we show that no $big(frac{e}{e-1} - Omega(1) big) Delta$-edge-coloring algorithm exists. We then provide an optimal, $big(frac{e}{e-1}+o(1)big)Delta$-edge-coloring algorithm for unknown $Delta=omega(log n)$. Key to our results, and of possible independent interest, is a novel fractional relaxation for edge coloring, for which we present optimal fractional online algorithms and a near-lossless online rounding scheme, yielding our optimal randomized algorithms.
The following online bin packing problem is considered: Items with integer sizes are given and variable sized bins arrive online. A bin must be used if there is still an item remaining which fits in it when the bin arrives. The goal is to minimize th
e total size of all the bins used. Previously, a lower bound of 5/4 on the competitive ratio of this problem was achieved using jobs of size S and 2S-1. For these item sizes and maximum bin size 4S-3, we obtain asymptotically matching upper and lower bounds, which vary depending on the ratio of the number of small jobs to the number of large jobs.
We consider the following online optimization problem. We are given a graph $G$ and each vertex of the graph is assigned to one of $ell$ servers, where servers have capacity $k$ and we assume that the graph has $ell cdot k$ vertices. Initially, $G$ d
oes not contain any edges and then the edges of $G$ are revealed one-by-one. The goal is to design an online algorithm $operatorname{ONL}$, which always places the connected components induced by the revealed edges on the same server and never exceeds the server capacities by more than $varepsilon k$ for constant $varepsilon>0$. Whenever $operatorname{ONL}$ learns about a new edge, the algorithm is allowed to move vertices from one server to another. Its objective is to minimize the number of vertex moves. More specifically, $operatorname{ONL}$ should minimize the competitive ratio: the total cost $operatorname{ONL}$ incurs compared to an optimal offline algorithm $operatorname{OPT}$. Our main contribution is a polynomial-time randomized algorithm, that is asymptotically optimal: we derive an upper bound of $O(log ell + log k)$ on its competitive ratio and show that no randomized online algorithm can achieve a competitive ratio of less than $Omega(log ell + log k)$. We also settle the open problem of the achievable competitive ratio by deterministic online algorithms, by deriving a competitive ratio of $Theta(ell lg k)$; to this end, we present an improved lower bound as well as a deterministic polynomial-time online algorithm. Our algorithms rely on a novel technique which combines efficient integer programming with a combinatorial approach for maintaining ILP solutions. We believe this technique is of independent interest and will find further applications in the future.
We consider a fundamental algorithmic question in spectral graph theory: Compute a spectral sparsifier of random-walk matrix-polynomial $$L_alpha(G)=D-sum_{r=1}^dalpha_rD(D^{-1}A)^r$$ where $A$ is the adjacency matrix of a weighted, undirected graph,
$D$ is the diagonal matrix of weighted degrees, and $alpha=(alpha_1...alpha_d)$ are nonnegative coefficients with $sum_{r=1}^dalpha_r=1$. Recall that $D^{-1}A$ is the transition matrix of random walks on the graph. The sparsification of $L_alpha(G)$ appears to be algorithmically challenging as the matrix power $(D^{-1}A)^r$ is defined by all paths of length $r$, whose precise calculation would be prohibitively expensive. In this paper, we develop the first nearly linear time algorithm for this sparsification problem: For any $G$ with $n$ vertices and $m$ edges, $d$ coefficients $alpha$, and $epsilon > 0$, our algorithm runs in time $O(d^2mlog^2n/epsilon^{2})$ to construct a Laplacian matrix $tilde{L}=D-tilde{A}$ with $O(nlog n/epsilon^{2})$ non-zeros such that $tilde{L}approx_{epsilon}L_alpha(G)$. Matrix polynomials arise in mathematical analysis of matrix functions as well as numerical solutions of matrix equations. Our work is particularly motivated by the algorithmic problems for speeding up the classic Newtons method in applications such as computing the inverse square-root of the precision matrix of a Gaussian random field, as well as computing the $q$th-root transition (for $qgeq1$) in a time-reversible Markov model. The key algorithmic step for both applications is the construction of a spectral sparsifier of a constant degree random-walk matrix-polynomials introduced by Newtons method. Our algorithm can also be used to build efficient data structures for effective resistances for multi-step time-reversible Markov models, and we anticipate that it could be useful for other tasks in network analysis.
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
