Results 11 to 20 of about 5,544,078 (251)

Homogeneous Linear Inequality Constraints for Neural Network Activations

open access: yes2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW), 2019
We propose a method to impose homogeneous linear inequality constraints of the form Ax ≤ 0 on neural network activations. The proposed method allows a data-driven training approach to be combined with modeling prior knowledge about the task.
Thomas Frerix, M. Nießner, D. Cremers
semanticscholar   +1 more source

Trust-region problems with linear inequality constraints: exact SDP relaxation, global optimality and robust optimization [PDF]

open access: yesMathematical programming, 2013
The trust-region problem, which minimizes a nonconvex quadratic function over a ball, is a key subproblem in trust-region methods for solving nonlinear optimization problems.
V. Jeyakumar, Guoyin Li
semanticscholar   +1 more source

Linear generalizations of Gronwall’s inequality [PDF]

open access: yesProceedings of the American Mathematical Society, 1976
A variety of linear generalizations of Gronwall's inequality, including recent multivariable results of D. R. Snow and E. C. Young, are subsumed and extended by simple arguments involving the resolvent kernel of the integral operator.
Jagdish Chandra, Paul Davis
openaire   +1 more source

An inequality for linear transformations [PDF]

open access: yesProceedings of the American Mathematical Society, 1967
1. Statements of results. In this paper the following elementary inequality is proved and exploited. THEOREM 1. If L is a positive-definite hermitian transformation on the finite dimensional unitary space V and p > 1, then for arbitrary vectors u and v (I) f|Uff2 + (L-PV, V) > ((I + L)-Pu + v, U + v). From (1) we can conclude THEOREM 2.
openaire   +2 more sources

Conditional Information Inequalities for Entropic and Almost Entropic Points [PDF]

open access: yesIEEE Transactions on Information Theory 59(11), 2013, pp. 7149-7167, 2012
We study conditional linear information inequalities, i.e., linear inequalities for Shannon entropy that hold for distributions whose entropies meet some linear constraints. We prove that some conditional information inequalities cannot be extended to any unconditional linear inequalities. Some of these conditional inequalities hold for almost entropic
arxiv   +1 more source

Stability of the Brascamp-Lieb constant and applications [PDF]

open access: yes, 2018
We prove that the best constant in the general Brascamp-Lieb inequality is a locally bounded function of the underlying linear transformations. As applications we deduce certain very general Fourier restriction, Kakeya-type, and nonlinear variants of the
Bennett, Jonathan   +3 more
core   +3 more sources

Analysis of junior high school students’ attempt to solve a linear inequality problem

open access: yes, 2017
Linear inequality is one of fundamental subjects within junior high school mathematics curricula. Several studies have been conducted to asses students’ perform on linear inequality.
M. Taqiyuddin, E. Sumiaty, A. Jupri
semanticscholar   +1 more source

Eigenvalue extensions of Bohr's inequality [PDF]

open access: yes, 2011
We present a weak majorization inequality and apply it to prove eigenvalue and unitarily invariant norm extensions of a version of the Bohr's inequality due to Vasi\'c and Ke\v{c}ki\'c.Comment: 8 pages, to appear in Linear Algebra ...
Jagjit Singh Matharu   +3 more
core   +2 more sources

How to Find New Characteristic-Dependent Linear Rank Inequalities using Secret Sharing [PDF]

open access: yesarXiv, 2021
Determining information ratios of access structures is an important problem in secret sharing. Information inequalities and linear rank inequalities play an important role for proving bounds. Characteristic-dependent linear rank inequalities are rank inequalities which are true over vector spaces with specific field characteristic. In this paper, using
arxiv  

A sharp nonlinear Hausdorff-Young inequality for small potentials [PDF]

open access: yes, 2018
The nonlinear Hausdorff-Young inequality follows from the work of Christ and Kiselev. Later Muscalu, Tao, and Thiele asked if the constants can be chosen independently of the exponent.
Kovač, Vjekoslav   +2 more
core   +2 more sources

Home - About - Disclaimer - Privacy