A Consistent Regularization Approach for Structured Prediction

TL;DR

This paper introduces a regularization method for structured prediction that embeds outputs in a linear space, ensuring consistency and providing theoretical and experimental validation of its effectiveness.

Contribution

It presents a novel regularization framework for structured prediction that guarantees universal consistency and offers finite sample generalization bounds.

Findings

Proposed approach is universally consistent.

Finite sample bounds are established.

Experimental results confirm practical usefulness.

Abstract

We propose and analyze a regularization approach for structured prediction problems. We characterize a large class of loss functions that allows to naturally embed structured outputs in a linear space. We exploit this fact to design learning algorithms using a surrogate loss approach and regularization techniques. We prove universal consistency and finite sample bounds characterizing the generalization properties of the proposed methods. Experimental results are provided to demonstrate the practical usefulness of the proposed approach.