Manifold Coordinates with Physical Meaning
arXiv:1811.11891
Abstract
Manifold embedding algorithms map high-dimensional data down to coordinates in a much lower-dimensional space. One of the aims of dimension reduction is to find intrinsic coordinates that describe the data manifold. The coordinates returned by the embedding algorithm are abstract, and finding their physical or domain-related meaning is not formalized and often left to domain experts. This paper studies the problem of recovering the meaning of the new low-dimensional representation in an automatic, principled fashion. We propose a method to explain embedding coordinates of a manifold as non-linear compositions of functions from a user-defined dictionary. We show that this problem can be set up as a sparse linear Group Lasso recovery problem, find sufficient recovery conditions, and demonstrate its effectiveness on data.
Submitted to JMLR. Improved over v2 (added appendix). Improved over v1 (revisions)
References in corpus (13)
- Discovering governing equations from data: Sparse identification of nonlinear dynamical systems
- Data-driven discovery of coordinates and governing equations
- Sanity Checks for Saliency Maps
- Graph Dynamical Networks for Unsupervised Learning of Atomic Scale Dynamics in Materials
- Between classical and quantum
- InfoGAN-CR and ModelCentrality: Self-supervised Model Training and Selection for Disentangling GANs
- Nonlinear Intrinsic Variables and State Reconstruction in Multiscale Simulations
- Non-linear dimensionality reduction: Riemannian metric estimation and the problem of geometric discovery
- Structure-adaptive manifold estimation
- Manifold Learning Using Kernel Density Estimation and Local Principal Components Analysis
- megaman: Manifold Learning with Millions of points
- Improved Support Recovery Guarantees for the Group Lasso With Applications to Structural Health Monitoring
- Parameterization of non-linear manifolds