MLLGJul 29, 2020

Learning Output Embeddings in Structured Prediction

arXiv:2007.14703v3
AI Analysis

This work addresses computational bottlenecks in structured prediction for machine learning practitioners, offering a more efficient and versatile tool for handling large datasets.

The paper tackles the computational complexity and flexibility of structured prediction by jointly learning a finite approximation of output embeddings and the regression function, resulting in a consistent estimator with an excess risk bound and significantly reduced computational complexity.

A powerful and flexible approach to structured prediction consists in embedding the structured objects to be predicted into a feature space of possibly infinite dimension by means of output kernels, and then, solving a regression problem in this output space. A prediction in the original space is computed by solving a pre-image problem. In such an approach, the embedding, linked to the target loss, is defined prior to the learning phase. In this work, we propose to jointly learn a finite approximation of the output embedding and the regression function into the new feature space. For that purpose, we leverage a priori information on the outputs and also unexploited unsupervised output data, which are both often available in structured prediction problems. We prove that the resulting structured predictor is a consistent estimator, and derive an excess risk bound. Moreover, the novel structured prediction tool enjoys a significantly smaller computational complexity than former output kernel methods. The approach empirically tested on various structured prediction problems reveals to be versatile and able to handle large datasets.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes