A pr 2 01 1 Tight bounds on the maximum size of a set of permutations with bounded VC - dimension ∗
نویسندگان
چکیده
The VC-dimension of a family P of n-permutations is the largest integer k such that the set of restrictions of the permutations in P on some k-tuple of positions is the set of all k! permutation patterns. Let rk(n) be the maximum size of a set of n-permutations with VC-dimension k. Raz showed that r2(n) grows exponentially in n. We show that r3(n) = 2 Θ(nα(n)) and for every t ≥ 1, we have r2t+2(n) = 2 Θ(nα(n)t) and r2t+3(n) = 2 O(nα(n)t logα(n)). We also study the maximum number pk(n) of 1-entries in an n×n (0, 1)-matrix with no (k+1)-tuple of columns containing all (k+1)-permutation matrices. We determine that p3(n) = Θ(nα(n)) and p2t+2(n) = n2 (1/t!)α(n)t±O(α(n)t−1) for every t ≥ 1. We also show that for every positive s there is a slowly growing function ζs(m) (for example ζs(m) = 2 O(α(s−3)/2(m)) for every odd s ≥ 5) satisfying the following. For all positive integers m,n,B and every m × n (0, 1)-matrix M with ζs(m)Bn 1-entries, the rows of M can be partitioned into s intervals so that some ⌊Bn/m⌋tuple of columns contains at least B 1-entries in each of the intervals.
منابع مشابه
Tight bounds on the maximum size of a set of permutations with bounded VC-dimension
The VC-dimension of a family P of n-permutations is the largest integer k such that the set of restrictions of the permutations in P on some k-tuple of positions is the set of all k! permutation patterns. Let rk(n) be the maximum size of a set of n-permutations with VCdimension k. Raz showed that r2(n) grows exponentially in n. We show that r3(n) = 2 Θ(n logα(n)) and for every t ≥ 1, we have r2...
متن کاملSet Systems and Families of Permutations with Small Traces
We study the maximum size of a set system on n elements whose trace on any b elements has size at most k. This question extends to hypergraphs the classical Dirac-type problems from extremal graph theory. We show that if for some b ≥ i ≥ 0 the shatter function fR of a set system ([n], R) satisfies fR(b) < 2 (b − i + 1) then |R| = O(n); this generalizes Sauer’s Lemma on the size of set systems w...
متن کاملepsilon-Samples of Kernels
We study the worst case error of kernel density estimates via subset approximation. A kernel density estimate of a distribution is the convolution of that distribution with a fixed kernel (e.g. Gaussian kernel). Given a subset (i.e. a point set) of the input distribution, we can compare the kernel density estimates of the input distribution with that of the subset and bound the worst case error...
متن کاملTight Bounds for the VC-Dimension of Piecewise Polynomial Networks
O(ws(s log d+log(dqh/ s))) and O(ws((h/ s) log q) +log(dqh/ s)) are upper bounds for the VC-dimension of a set of neural networks of units with piecewise polynomial activation functions, where s is the depth of the network, h is the number of hidden units, w is the number of adjustable parameters, q is the maximum of the number of polynomial segments of the activation function, and d is the max...
متن کاملThe VC-dimension of graphs with respect to k-connected subgraphs
We study the VC-dimension of the set system on the vertex set of some graph which is induced by the family of its k-connected subgraphs. In particular, we give tight upper and lower bounds for the VC-dimension. Moreover, we show that computing the VC-dimension is NP-complete and that it remains NP-complete for split graphs and for some subclasses of planar bipartite graphs in the cases k = 1 an...
متن کامل