no code implementations • 13 Dec 2020 • Vincent Corlay, Joseph J. Boutros, Philippe Ciblat, Loïc Brunel
It is exponential in the space dimension $n$, which induces shallow neural networks of exponential size.
no code implementations • 28 Feb 2019 • Vincent Corlay, Joseph J. Boutros, Philippe Ciblat, Loic Brunel
We prove that they can be computed by ReLU networks with quadratic depth and linear width in the space dimension.
no code implementations • 6 Feb 2019 • Vincent Corlay, Joseph J. Boutros, Philippe Ciblat, Loic Brunel
Lattice decoding in Rn, known as the closest vector problem (CVP), becomes a classification problem in the fundamental parallelotope with a piecewise linear function defining the boundary.
no code implementations • 4 Dec 2018 • Vincent Corlay, Joseph J. Boutros, Philippe Ciblat, Loïc Brunel
A quasi-static flat multiple-antenna channel is considered.