On the CVP for the root lattices via folding with deep ReLU neural networks

Abstract : Point lattices and their decoding via neural networks are considered in this paper. Lattice decoding in R n , known as the closest vector problem (CVP), becomes a classification problem in the fundamental parallelotope with a piecewise linear function defining the boundary. Theoretical results are obtained by studying root lattices. We show how the number of pieces in the boundary function reduces dramatically with folding, from exponential to linear. This translates into a two-layer ReLU network requiring a number of neurons growing exponentially in n to solve the CVP, whereas this complexity becomes polynomial in n for a deep ReLU network.
Complete list of metadatas

Cited literature [15 references]  Display  Hide  Download

Contributor : Philippe Ciblat <>
Submitted on : Monday, September 9, 2019 - 6:16:33 PM
Last modification on : Wednesday, September 11, 2019 - 1:28:18 AM


Files produced by the author(s)


  • HAL Id : hal-02282190, version 1


Vincent Corlay, Joseph Boutros, Philippe Ciblat, Loïc Brunel. On the CVP for the root lattices via folding with deep ReLU neural networks. IEEE International Symposium on Information Thoery (ISIT), Jul 2019, Paris, France. ⟨hal-02282190⟩



Record views


Files downloads