Large-scale semi-supervised learning with online spectral graph sparsification

TL;DR

The paper presents Sparse-HFS, a scalable semi-supervised learning algorithm that efficiently solves SSL problems with minimal space and time complexity, suitable for large datasets.

Contribution

It introduces Sparse-HFS, a novel algorithm that significantly reduces computational resources needed for large-scale semi-supervised learning.

Findings

Uses only O(n polylog(n)) space

Operates in O(m polylog(n)) time

Enables scalable SSL solutions for large datasets

Abstract

We introduce Sparse-HFS, a scalable algorithm that can compute solutions to SSL problems using only O(n polylog(n)) space and O(m polylog(n)) time.