How fast can a threshold gate learn?
Proceedings of a workshop on Computational learning theory and natural learning systems (vol. 1) : constraints and prospects: constraints and prospects
Learning with restricted focus of attention
Journal of Computer and System Sciences
Robust Solutions of Uncertain Quadratic and Conic-Quadratic Problems
SIAM Journal on Optimization
Every Linear Threshold Function has a Low-Weight Approximator
Computational Complexity
A faster PSPACE algorithm for deciding the existential theory of the reals
SFCS '88 Proceedings of the 29th Annual Symposium on Foundations of Computer Science
Semidefinite Relaxation Bounds for Indefinite Homogeneous Quadratic Optimization
SIAM Journal on Optimization
Improved approximation bound for quadratic optimization problems with orthogonality constraints
SODA '09 Proceedings of the twentieth Annual ACM-SIAM Symposium on Discrete Algorithms
Improved Approximation of Linear Threshold Functions
CCC '09 Proceedings of the 2009 24th Annual IEEE Conference on Computational Complexity
SIAM Journal on Computing
Bounded Independence Fools Halfspaces
SIAM Journal on Computing
SIAM Journal on Computing
COLT'06 Proceedings of the 19th annual conference on Learning Theory
Nonnegative k-sums, fractional covers, and probability of small deviations
Journal of Combinatorial Theory Series B
Nearly optimal solutions for the chow parameters problem and low-weight approximation of halfspaces
STOC '12 Proceedings of the forty-fourth annual ACM symposium on Theory of computing
Hi-index | 0.00 |
This paper makes two contributions towards determining some well-studied optimal constants in Fourier analysis of Boolean functions and high-dimensional geometry. 1 It has been known since 1994 [GL94] that every linear threshold function has squared Fourier mass at least 1/2 on its degree-0 and degree-1 coefficients. Denote the minimum such Fourier mass by W≤1[LTF], where the minimum is taken over all n-variable linear threshold functions and all n≥0. Benjamini, Kalai and Schramm [BKS99] have conjectured that the true value of W≤1[LTF] is 2/π. We make progress on this conjecture by proving that W≤1[LTF]≥1/2+c for some absolute constant c0. The key ingredient in our proof is a "robust" version of the well-known Khintchine inequality in functional analysis, which we believe may be of independent interest. 2 We give an algorithm with the following property: given any η0, the algorithm runs in time 2poly(1/η) and determines the value of W≤1[LTF] up to an additive error of ±η. We give a similar 2poly(1/η)-time algorithm to determine Tomaszewski's constant to within an additive error of ±η; this is the minimum (over all origin-centered hyperplanes H) fraction of points in {−1,1}n that lie within Euclidean distance 1 of H. Tomaszewski's constant is conjectured to be 1/2; lower bounds on it have been given by Holzman and Kleitman [HK92] and independently by Ben-Tal, Nemirovski and Roos [BTNR02]. Our algorithms combine tools from anti-concentration of sums of independent random variables, Fourier analysis, and Hermite analysis of linear threshold functions.