Analyzing the Benefits of Prototypes for Semi-Supervised Category Learning
This work addresses semi-supervised learning for category formation, offering an incremental improvement by applying prototype-based priors in variational auto-encoders.
The paper tackles the problem of semi-supervised category learning by examining prototype-based representations in a Bayesian unsupervised learning model, showing that forming prototypes improves performance on image datasets and leads to clustered latent embeddings.
Categories can be represented at different levels of abstraction, from prototypes focused on the most typical members to remembering all observed exemplars of the category. These representations have been explored in the context of supervised learning, where stimuli are presented with known category labels. We examine the benefits of prototype-based representations in a less-studied domain: semi-supervised learning, where agents must form unsupervised representations of stimuli before receiving category labels. We study this problem in a Bayesian unsupervised learning model called a variational auto-encoder, and we draw on recent advances in machine learning to implement a prior that encourages the model to use abstract prototypes to represent data. We apply this approach to image datasets and show that forming prototypes can improve semi-supervised category learning. Additionally, we study the latent embeddings of the models and show that these prototypes allow the models to form clustered representations without supervision, contributing to their success in downstream categorization performance.