Ridge Regression, Hubness, and Zero-Shot Learning

TL;DR

This paper proposes a novel approach to zero-shot learning by mapping labels into the example space using ridge regression, which reduces hubness and improves accuracy in tasks like bilingual lexicon extraction and image labeling.

Contribution

The paper introduces a new method of mapping labels into the example space to suppress hubness, contrasting with traditional methods that map examples to labels.

Findings

Hubness was reduced in both bilingual lexicon extraction and image labeling tasks.

Mapping labels into the example space improves zero-shot learning accuracy.

Theoretical proof supports the effectiveness of the proposed approach.

Abstract

This paper discusses the effect of hubness in zero-shot learning, when ridge regression is used to find a mapping between the example space to the label space. Contrary to the existing approach, which attempts to find a mapping from the example space to the label space, we show that mapping labels into the example space is desirable to suppress the emergence of hubs in the subsequent nearest neighbor search step. Assuming a simple data model, we prove that the proposed approach indeed reduces hubness. This was verified empirically on the tasks of bilingual lexicon extraction and image labeling: hubness was reduced with both of these tasks and the accuracy was improved accordingly.