Constructive Discrepancy Minimization for Convex Sets

  title={Constructive Discrepancy Minimization for Convex Sets},
  author={Thomas Rothvoss},
  journal={2014 IEEE 55th Annual Symposium on Foundations of Computer Science},
  • T. Rothvoss
  • Published 1 April 2014
  • Mathematics
  • 2014 IEEE 55th Annual Symposium on Foundations of Computer Science
A classical theorem of Spencer shows that any set system with n sets and n elements admits a coloring of discrepancy O(√(n)). Recent exciting work of Bansal, Lovett and Meka shows that such colorings can be found in polynomial time. In fact, the Lovett-Meka algorithm finds a half integral point in any "large enough" polytope. However, their algorithm crucially relies on the facet structure and does not apply to general convex sets. We show that for any symmetric convex set K with measure at… 

Figures from this paper

Efficient algorithms for discrepancy minimization in convex sets
A constructive version of the result of Gluskin and Giannopoulos, in which the coloring is attained via the optimization of a linear function is proved, which implies a linear programming based algorithm for combinatorial discrepancy obtaining the same result as Spencer.
Algorithms for combinatorial discrepancy
  • S. Garg
  • Mathematics, Computer Science
  • 2018
This thesis first attempted to solve the problem of discrepancy minimization using a convex geometric approach which led to improved algorithms in some regimes, and developed an algorithm for the most important special case when the convex body K is a scaling of the hypercube.
Improved Algorithmic Bounds for Discrepancy of Sparse Set Systems
An algorithm is given that finds a coloring with discrepancy $O((t \log n \log s)^{1/2})$ where $s$ is the maximum cardinality of a set, which improves upon the previous constructive bound of $O(t^2)$ based on algorithmic variants of the partial coloring method.
2 Combinatorial Discrepancy for Boxes via the γ 2 Norm
A new lower bound of Ω(logd−1 n) for the d-dimensional Tusnády problem, asking for the combinatorial discrepancy of an n-point set in R with respect to axis-parallel boxes is proved and a new, very simple proof is obtained.
A size-sensitive discrepancy bound for set systems of bounded primal shatter dimension
Let (X, S) be a set system on an n-point set X. The discrepancy of S is defined as the minimum of the largest deviation from an even split, over all subsets of S e S and two-colorings χ on X. We
Algorithms in Discrepancy Theory and Lattices
Algorithms in Discrepancy Theory and Lattices Harishchandra Ramadas Chair of the Supervisory Committee: Assistant Professor Thomas Rothvoß Department of Mathematics This thesis deals with algorithmic
Hyperbolic Concentration, Anti-concentration, and Discrepancy
This paper presents the first work that shows either concentration or anti-concentration results for hyperbolic polynomials and implies that a random bi-coloring of any set system with n sets and n elements will have discrepancy O ( √ n log n ) with high probability.
Report-Analytical Methods in TCS Constructive Discrepancy Minimization for Convex Sets
This report relates the main results and methods of Constructive Discrepancy Minimization for Convex Sets [6] by Thomas Rothvoß, a beautiful and largely self-contained paper in algorithmic
Approximating Hereditary Discrepancy via Small Width Ellipsoids
A direct simple O(log3/2 n)-approximation algorithm is given that up to this approximation factor, the hereditary discrepancy of a matrix A is characterized by the optimal value of simple geometric convex program that seeks to minimize the largest l∞ norm of any point in a ellipsoid containing the columns of A.
Ju l 2 01 4 Approximating Hereditary Discrepancy via Small Width Ellipsoids
The Discrepancy of a hypergraph is the minimum attainable value, over twocolorings of its vertices, of the maximum absolute imbalance of any hyperedge. The Hereditary Discrepancy of a hypergraph,


Constructive Discrepancy Minimization by Walking on the Edges
A new randomized algorithm to find a coloring as in Spencer's result based on a restricted random walk which is “truly” constructive in that it does not appeal to the existential arguments, giving a new proof of Spencer's theorem and the partial coloring lemma.
Improving the discrepancy bound for sparse matrices: better approximations for sparse lattice approximation problems
Improved upper bounds on the discrepancy of two well-studied families of sparse matrices: e permutations of [n], and rectangles containing n points in Rk are shown, and a discrepancy bound of O(&logn) is shown for the former, improving on the previous-best O( logn) due to Bohus.
The ellipsoid method and its consequences in combinatorial optimization
The method yields polynomial algorithms for vertex packing in perfect graphs, for the matching and matroid intersection problems, for optimum covering of directed cuts of a digraph, and for the minimum value of a submodular set function.
Beck's Three Permutations Conjecture: A Counterexample and Some Consequences
This work constructs three permutations whose corresponding set system has discrepancy Ω(log n), and proves an interesting implication of this construction in the reverse direction: there are instances of bin packing and corresponding optimal basic feasible solutions for the Gilmore-Gomory LP relaxation such that any packing that contains only patterns from the support of these solutions requires at least opt + Ω (log m) bins.
The discrepancy of permutation families
In this note, we show that the discrepancy of any family of ‘ permutations of [n] = f1; 2;:::;ng is O( p ‘ logn), improving on the O(‘ logn) bound due to Bohus (Random Structures & Algorithms,
Interlacing families II: Mixed characteristic polynomials and the Kadison{Singer problem
We use the method of interlacing polynomials introduced in our previous article to prove two theorems known to imply a positive solution to the Kadison{Singer problem. The rst is Weaver’s conjecture
On the Discrepancy of 3 Permutations
  • G. Bohus
  • Mathematics
    Random Struct. Algorithms
  • 1990
It is shown that for any constant number of orderings the discrepancy is O(log n) and the proof also gives an efficient algorithm to determine such a coloring.
The discrepancy method - randomness and complexity
This book tells the story of the discrepancy method in a few short independent vignettes. It is a varied tale which includes such topics as communication complexity, pseudo-randomness, rapidly mixing
Probability in Banach Spaces: Isoperimetry and Processes
Notation.- 0. Isoperimetric Background and Generalities.- 1. Isoperimetric Inequalities and the Concentration of Measure Phenomenon.- 2. Generalities on Banach Space Valued Random Variables and
It is proved that the distribution function for the maximum of the modulus of a set of jointly Gaussian random variables with given variance and zero mean is minimal if these variables are