Augment and Reduce: Stochastic Inference for Large Categorical Distributions
This addresses a computational bottleneck for researchers and practitioners working with large categorical models in areas like classification and language models, though it is an incremental improvement over existing methods.
The paper tackles the computational expense of large categorical distributions in machine learning by proposing the augment and reduce (A&R) method, which uses latent variable augmentation and stochastic variational inference to maximize a lower bound on marginal likelihood, resulting in a tighter bound and better predictive performance on large-scale classification problems.
Categorical distributions are ubiquitous in machine learning, e.g., in classification, language models, and recommendation systems. However, when the number of possible outcomes is very large, using categorical distributions becomes computationally expensive, as the complexity scales linearly with the number of outcomes. To address this problem, we propose augment and reduce (A&R), a method to alleviate the computational complexity. A&R uses two ideas: latent variable augmentation and stochastic variational inference. It maximizes a lower bound on the marginal likelihood of the data. Unlike existing methods which are specific to softmax, A&R is more general and is amenable to other categorical models, such as multinomial probit. On several large-scale classification problems, we show that A&R provides a tighter bound on the marginal likelihood and has better predictive performance than existing approaches.