DeepAI AI Chat
Log In Sign Up

Algorithms for metric learning via contrastive embeddings

by   Diego Ihara Centurion, et al.
University of Illinois at Chicago

We study the problem of supervised learning a metric space under discriminative constraints. Given a universe X and sets S, D⊂X 2 of similar and dissimilar pairs, we seek to find a mapping f:X→ Y, into some target metric space M=(Y,ρ), such that similar objects are mapped to points at distance at most u, and dissimilar objects are mapped to points at distance at least ℓ. More generally, the goal is to find a mapping of maximum accuracy (that is, fraction of correctly classified pairs). We propose approximation algorithms for various versions of this problem, for the cases of Euclidean and tree metric spaces. For both of these target spaces, we obtain fully polynomial-time approximation schemes (FPTAS) for the case of perfect information. In the presence of imperfect information we present approximation algorithms that run in quasipolynomial time (QPTAS). Our algorithms use a combination of tools from metric embeddings and graph partitioning, that could be of independent interest.


page 1

page 2

page 3

page 4


Approximation Algorithms For The Dispersion Problems in a Metric Space

In this article, we consider the c-dispersion problem in a metric space ...

Learning Mahalanobis Metric Spaces via Geometric Approximation Algorithms

Learning Mahalanobis metric spaces is an important problem that has foun...

MSPP: A Highly Efficient and Scalable Algorithm for Mining Similar Pairs of Points

The closest pair of points problem or closest pair problem (CPP) is an i...

Distant Representatives for Rectangles in the Plane

The input to the distant representatives problem is a set of n objects i...

Comparing Distributions and Shapes using the Kernel Distance

Starting with a similarity function between objects, it is possible to d...

Negative type diversities, a multi-dimensional analogue of negative type metrics

Diversities are a generalization of metric spaces in which a non-negativ...

Generalized Metric Repair on Graphs

Many modern data analysis algorithms either assume that or are considera...