Contrastive Similarity Matching for Supervised Learning

TL;DR

This paper introduces a biologically-plausible deep learning method that aligns layer-specific similarity matrices with observed neural and visual pathway behaviors, using a contrastive similarity matching objective.

Contribution

It presents a novel contrastive similarity matching framework with biologically plausible neural dynamics and learning rules for supervised deep learning.

Findings

Derives deep neural networks with biologically plausible plasticity rules.

Proposes a layer-specific similarity learning objective.

Connects the method to energy-based learning algorithms.

Abstract

We propose a novel biologically-plausible solution to the credit assignment problem motivated by observations in the ventral visual pathway and trained deep neural networks. In both, representations of objects in the same category become progressively more similar, while objects belonging to different categories become less similar. We use this observation to motivate a layer-specific learning goal in a deep network: each layer aims to learn a representational similarity matrix that interpolates between previous and later layers. We formulate this idea using a contrastive similarity matching objective function and derive from it deep neural networks with feedforward, lateral, and feedback connections, and neurons that exhibit biologically-plausible Hebbian and anti-Hebbian plasticity. Contrastive similarity matching can be interpreted as an energy-based learning algorithm, but with significant differences from others in how a contrastive function is constructed.