Model-centric Data Manifold: the Data Through the Eyes of the Model

DOI10.1137/21M1437056zbMATH Open1523.68079arXiv2104.13289MaRDI QIDQ6366265FDOQ6366265

Publication date: 26 April 2021

Abstract: We discover that deep ReLU neural network classifiers can see a low-dimensional Riemannian manifold structure on data. Such structure comes via the local data matrix, a variation of the Fisher information matrix, where the role of the model parameters is taken by the data variables. We obtain a foliation of the data domain and we show that the dataset on which the model is trained lies on a leaf, the data leaf, whose dimension is bounded by the number of classification labels. We validate our results with some experiments with the MNIST dataset: paths on the data leaf connect valid images, while other leaves cover noisy images.

Mathematics Subject Classification ID

Artificial neural networks and deep learning (68T07) Computational aspects of data analysis and big data (68T09) Local Riemannian geometry (53B20) Differential geometric aspects of statistical manifolds and information geometry (53B12)

This page was built for publication: Model-centric Data Manifold: the Data Through the Eyes of the Model

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6366265)