Estimating the Euclidean distortion of an orbit space
By: Ben Blum-Smith , Harm Derksen , Dustin G. Mixon and more
Potential Business Impact:
Makes math easier for smart computer learning.
Given a finite-dimensional inner product space $V$ and a group $G$ of isometries, we consider the problem of embedding the orbit space $V/G$ into a Hilbert space in a way that preserves the quotient metric as well as possible. This inquiry is motivated by applications to invariant machine learning. We introduce several new theoretical tools before using them to tackle various fundamental instances of this problem.
Similar Papers
Finite sample bounds for barycenter estimation in geodesic spaces
Statistics Theory
Finds the average point in curved spaces.
When Distances Lie: Euclidean Embeddings in the Presence of Outliers and Distance Violations
Computational Geometry
Fixes messy distance data for computer mapping.
A group-theoretic framework for machine learning in hyperbolic spaces
Machine Learning (CS)
Makes computer learning faster and smarter.