arXiv · 2306.17418
ReLU Neural Networks, Polyhedral Decompositions, and Persistent Homolog
Abstract
A ReLU neural network leads to a finite polyhedral decomposition of input space and a corresponding finite dual graph. We show that while this dual graph is a coarse quantization of input space, it is sufficiently robust that it can be combined with persistent homology to detect homological signals of manifolds in the input space from samples. This property holds for a variety of networks trained for a wide range of purposes that have nothing to do with this topological application. We found this feature to be surprising and interesting; we hope it will also be useful.
Explore related subjects
Keep this discovery
Explore connections, maps & timelines
Yajing Liu, Christina M Cole, Chris Peterson, Michael Kirby. 2023-06-30. ReLU Neural Networks, Polyhedral Decompositions, and Persistent Homolog. https://arxiv.org/abs/2306.17418
Cite the original work for its findings. Save a collection to share your selection of sources.