Semi-supervised learning with Bayesian Confidence Propagation Neural Network

TL;DR

This paper explores how Bayesian Confidence Propagation Neural Networks can learn useful representations from limited labeled data and compares their semi-supervised classification performance with other methods.

Contribution

It introduces a semi-supervised learning approach using BCPNN and evaluates its effectiveness against other classifiers.

Findings

BCPNN effectively learns representations with few labels

Semi-supervised BCPNN outperforms some traditional classifiers

The approach demonstrates biological plausibility and competitive accuracy

Abstract

Learning internal representations from data using no or few labels is useful for machine learning research, as it allows using massive amounts of unlabeled data. In this work, we use the Bayesian Confidence Propagation Neural Network (BCPNN) model developed as a biologically plausible model of the cortex. Recent work has demonstrated that these networks can learn useful internal representations from data using local Bayesian-Hebbian learning rules. In this work, we show how such representations can be leveraged in a semi-supervised setting by introducing and comparing different classifiers. We also evaluate and compare such networks with other popular semi-supervised classifiers.