Searcharxiv⌕ Search

arXiv subjects

Joel Spencer

Publications and source records attributed to Joel Spencer.

29 records · Page 2Linked to original sources
arXiv2004

Simulating a Random Walk with Constant Error

We analyze Jim Propp's P-machine, a simple deterministic process that simulates a random walk on $Z^d$ to within a constant. The proof of the error bound relies on several estimates in the theory of simple random walks and some careful summing. We mention three intriguing conjectures concerning sign-changes and unimodality of functions in the linear span of $\{p(\cdot,x) : x \in Z^d\}$, where $p(n,x)$ is the probability that a walk beginning from the origin arrives at $x$ at time $n$.

math.CO↗
arXiv2004

Succinct Definitions in the First Order Theory of Graphs

We say that a first order sentence A defines a graph G if A is true on G but false on any graph non-isomorphic to G. Let L(G) (resp. D(G)) denote the minimum length (resp. quantifier rank) of a such sentence. We define the succinctness function s(n) (resp. its variant q(n)) to be the minimum L(G) (resp. D(G)) over all graphs on n vertices. We prove that s(n) and q(n) may be so small that for no general recursive function f we can have f(s(n))\ge n for all n. However, for the function q^*(n)=\max_{i\le n}q(i), which is the least monotone nondecreasing function bounding q(n) from above, we have q^*(n)=(1+o(1))\log^*n, where \log^*n equals the minimum number of iterations of the binary logarithm sufficient to lower n below 1. We show an upper bound q(n)<\log^*n+5 even under the restriction of the class of graphs to trees. Under this restriction, for q(n) we also have a matching lower bound. We show a relationship D(G)\ge(1-o(1))\log^*L(G) and prove, using the upper bound for q(n), that this relationship is tight. For a non-negative integer a, let D_a(G) and q_a(n) denote the analogs of D(G) and q(n) for defining formulas in the negation normal form with at most a quantifier alternations in any sequence of nested quantifiers. We show a superrecursive gap between D_0(G) and D_3(G) and hence between D_0(G) and D(G). Despite it, for q_0(n) we still have a kind of log-star upper bound: q_0(n)\le2\log^*n+O(1) for infinitely many n.

math.LO↗
arXiv2004

How Complex are Random Graphs in First Order Logic?

It is not hard to write a first order formula which is true for a given graph G but is false for any graph not isomorphic to G. The smallest number $(G) of nested quantifiers in a such formula can serve as a measure for the ``first order complexity'' of G. Here, this parameter is studied for random graphs. We determine it asymptotically when the edge probability p is constant; in fact, D(G) is of order log n then. For very sparse graphs its magnitude is Θ(n). On the other hand, for certain (carefully chosen) values of p the parameter D(G) can drop down to the very slow growing function log^* n, the inverse of the tower-function. The general picture, however, is still a mystery.

math.CO↗
arXiv2004

Random subgraphs of finite graphs: I. The scaling window under the triangle condition

We study random subgraphs of an arbitrary finite connected transitive graph $\mathbb G$ obtained by independently deleting edges with probability $1-p$. Let $V$ be the number of vertices in $\mathbb G$, and let $Ω$ be their degree. We define the critical threshold $p_c=p_c(\mathbb G,λ)$ to be the value of $p$ for which the expected cluster size of a fixed vertex attains the value $λV^{1/3}$, where $λ$ is fixed and positive. We show that for any such model, there is a phase transition at $p_c$ analogous to the phase transition for the random graph, provided that a quantity called the triangle diagram is sufficiently small at the threshold $p_c$. In particular, we show that the largest cluster inside a scaling window of size $|p-p_c|=Θ(\cn^{-1}V^{-1/3})$ is of size $Θ(V^{2/3})$, while below this scaling window, it is much smaller, of order $O(ε^{-2}\log(Vε^3))$, with $ε=\cn(p_c-p)$. We also obtain an upper bound $O(\cn(p-p_c)V)$ for the expected size of the largest cluster above the window. In addition, we define and analyze the percolation probability above the window and show that it is of order $Θ(\cn(p-p_c))$. Among the models for which the triangle diagram is small enough to allow us to draw these conclusions are the random graph, the $n$-cube and certain Hamming cubes, as well as the spread-out $n$-dimensional torus for $n>6$.

math.PR↗
arXiv2004

Random subgraphs of finite graphs: II. The lace expansion and the triangle condition

In a previous paper, we defined a version of the percolation triangle condition that is suitable for the analysis of bond percolation on a finite connected transitive graph, and showed that this triangle condition implies that the percolation phase transition has many features in common with the phase transition on the complete graph. In this paper, we use a new and simplified approach to the lace expansion to prove quite generally that for finite graphs that are tori the triangle condition for percolation is implied by a certain triangle condition for simple random walks on the graph. The latter is readily verified for several graphs with vertex set $\{0,1,..., r-1\}^n$, including the Hamming cube on an alphabet of $r$ letters (the $n$-cube, for $r=2$), the $n$-dimensional torus with nearest-neighbor bonds and $n$ sufficiently large, and the $n$-dimensional torus with $n>6$ and sufficiently spread-out (long range) bonds. The conclusions of our previous paper thus apply to the percolation phase transition for each of the above examples.

math.PR↗
arXiv2004

Random subgraphs of finite graphs: III. The phase transition for the $n$-cube

We study random subgraphs of the $n$-cube $\{0,1\}^n$, where nearest-neighbor edges are occupied with probability $p$. Let $p_c(n)$ be the value of $p$ for which the expected cluster size of a fixed vertex attains the value $λ2^{n/3}$, where $λ$ is a small positive constant. Let $ε=n(p-p_c(n))$. In two previous papers, we showed that the largest cluster inside a scaling window given by $|ε|=Θ(2^{-n/3})$ is of size $Θ(2^{2n/3})$, below this scaling window it is at most $2(\log2) nε^{-2}$, and above this scaling window it is at most $O(ε2^n)$. In this paper, we prove that for $p - p_c(n) \geq e^{-cn^{1/3}}$ the size of the largest cluster is at least $Θ(ε2^n)$, which is of the same order as the upper bound. This provides an understanding of the phase transition that goes far beyond that obtained by previous authors. The proof is based on a method that has come to be known as ``sprinkling,'' and relies heavily on the specific geometry of the $n$-cube.

math.PR↗
arXiv1998

Packing Ferrers Shapes

Answering a question of Wilf, we show that if $n$ is sufficiently large, then one cannot cover an $n \times p(n)$ rectangle using each of the $p(n)$ distinct Ferrers shapes of size $n$ exactly once. Moreover, the maximum number of pairwise distinct, non-overlapping Ferrers shapes that can be packed in such a rectangle is only $Θ(p(n)/ \log n).$

math.CO↗
arXiv1995

Asymptotically optimal covering designs

A (v,k,t) covering design, or covering, is a family of k-subsets, called blocks, chosen from a v-set, such that each t-subset is contained in at least one of the blocks. The number of blocks is the covering's size}, and the minimum size of such a covering is denoted by C(v,k,t). It is easy to see that a covering must contain at least (v choose t)/(k choose t) blocks, and in 1985 Rödl [European J. Combin. 5 (1985), 69-78] proved a long-standing conjecture of Erdős and Hanani [Publ. Math. Debrecen 10 (1963), 10-13] that for fixed k and t, coverings of size (v choose t)/(k choose t) (1+o(1)) exist (as v \to \infty). An earlier paper by the first three authors [J. Combin. Des. 3 (1995), 269-284] gave new methods for constructing good coverings, and gave tables of upper bounds on C(v,k,t) for small v, k, and t. The present paper shows that two of those constructions are asymptotically optimal: For fixed k and t, the size of the coverings constructed matches Rödl's bound. The paper also makes the o(1) error bound explicit, and gives some evidence for a much stronger bound.

math.CO↗
arXiv1994

Can you feel the double jump?

Paul Erdős and Alfred Renyi considered the evolution of the random graph G(n,p) as p ``evolved'' from 0 to 1. At p=1/n a sudden and dramatic change takes place in G. When p=c/n with c<1 the random G consists of small components, the largest of size Theta(log n). But by p=c/n with c>1 many of the components have ``congealed'' into a ``giant component'' of size Theta (n). Erdős and Renyi called this the double jump, the terms phase transition (from the analogy to percolation) and Big Bang have also been proferred. Now imagine an observer who can only see G through a logical fog. He may refer to graph theoretic properties A within a limited logical language. Will he be able to detect the double jump? The answer depends on the strength of the language. Our rough answer to this rough question is: the double jump is not detectible in the First Order Theory of Graphs but it is detectible in the Second Order Monadic Theory of Graphs.

math.LO↗
arXiv1994

Random sparse unary predicates

The main result is the following Theorem: Let p=p(n) be such that p(n) in [0,1] for all n and either p(n)<< n^{-1} or for some positive integer k, n^{-1/k}<< p(n)<< n^{-1/(k+1)} or for all epsilon >0, n^{- epsilon}<< p(n) and n^{- epsilon}<< 1-p(n) or for some positive integer k, n^{-1/k}<< 1-p(n)<< n^{-1/(k+1)} or 1-p(n)<< n^{-1}. Then p(n) satisfies the Zero-One Law for circular unary predicates. Inversely, if p(n) falls into none of the above categories then it does not satisfy the Zero-One Law for circular unary predicates.

math.LO↗