Conditional Restricted Boltzmann Machines for Structured Output Prediction

TL;DR

This paper introduces improved training algorithms for Conditional Restricted Boltzmann Machines (CRBMs) tailored for structured output prediction tasks, addressing limitations of traditional methods and demonstrating superior performance.

Contribution

The paper proposes novel learning algorithms for CRBMs that outperform Contrastive Divergence in structured output prediction problems with different output space complexities.

Findings

New algorithms outperform Contrastive Divergence in experiments.

Effective for both small and large structured output spaces.

Applicable to tasks like multi-label classification and image denoising.

Abstract

Conditional Restricted Boltzmann Machines (CRBMs) are rich probabilistic models that have recently been applied to a wide range of problems, including collaborative filtering, classification, and modeling motion capture data. While much progress has been made in training non-conditional RBMs, these algorithms are not applicable to conditional models and there has been almost no work on training and generating predictions from conditional RBMs for structured output problems. We first argue that standard Contrastive Divergence-based learning may not be suitable for training CRBMs. We then identify two distinct types of structured output prediction problems and propose an improved learning algorithm for each. The first problem type is one where the output space has arbitrary structure but the set of likely output configurations is relatively small, such as in multi-label classification. The second problem is one where the output space is arbitrarily structured but where the output space variability is much greater, such as in image denoising or pixel labeling. We show that the new learning algorithms can work much better than Contrastive Divergence on both types of problems.