Metric learning approach for graph-based label propagation

TL;DR

This paper introduces a metric learning algorithm designed to optimize the vectorial representation of instances, thereby improving the construction of graphs used in semi-supervised label propagation tasks.

Contribution

It presents a novel metric learning approach that adapts the vector space to enhance graph-based semi-supervised learning performance.

Findings

Improved graph construction for label propagation

Enhanced semi-supervised learning accuracy

Adaptive metric learning outperforms Euclidean norm

Abstract

The efficiency of graph-based semi-supervised algorithms depends on the graph of instances on which they are applied. The instances are often in a vectorial form before a graph linking them is built. The construction of the graph relies on a metric over the vectorial space that help define the weight of the connection between entities. The classic choice for this metric is usually a distance measure or a similarity measure based on the euclidean norm. We claim that in some cases the euclidean norm on the initial vectorial space might not be the more appropriate to solve the task efficiently. We propose an algorithm that aims at learning the most appropriate vectorial representation for building a graph on which the task at hand is solved efficiently.