SqueezeFit: Label-aware dimensionality reduction by semidefinite programming

12/06/2018
by   Culver McWhirter, et al.
0

Given labeled points in a high-dimensional vector space, we seek a low-dimensional subspace such that projecting onto this subspace maintains some prescribed distance between points of differing labels. Intended applications include compressive classification. Taking inspiration from large margin nearest neighbor classification, this paper introduces a semidefinite relaxation of this problem. Unlike its predecessors, this relaxation is amenable to theoretical analysis, allowing us to provably recover a planted projection operator from the data.

READ FULL TEXT
research
11/13/2019

Topological Stability: Guided Determination of the Nearest Neighbors in Non-Linear Dimensionality Reduction Techniques

In machine learning field, dimensionality reduction is one of the import...
research
11/13/2019

Topological Stability: a New Algorithm for Selecting The Nearest Neighbors in Non-Linear Dimensionality Reduction Techniques

In the machine learning field, dimensionality reduction is an important ...
research
02/16/2017

RIPML: A Restricted Isometry Property based Approach to Multilabel Learning

The multilabel learning problem with large number of labels, features, a...
research
07/19/2011

Unsupervised K-Nearest Neighbor Regression

In many scientific disciplines structures in high-dimensional data have ...
research
05/25/2021

Hierarchical Subspace Learning for Dimensionality Reduction to Improve Classification Accuracy in Large Data Sets

Manifold learning is used for dimensionality reduction, with the goal of...
research
04/21/2019

A convex relaxation to compute the nearest structured rank deficient matrix

Given an affine space of matrices L and a matrix θ∈ L, consider the prob...
research
06/01/2022

Accelerated first-order methods for a class of semidefinite programs

This paper introduces a new storage-optimal first-order method (FOM), Ce...

Please sign up or login with your details

Forgot password? Click here to reset