Neuromorphic Deep Learning Machines

TL;DR

This paper introduces an event-driven random backpropagation (eRBP) rule for neuromorphic deep learning, enabling efficient, hardware-compatible learning of deep representations with high accuracy and robustness.

Contribution

It presents a novel, hardware-friendly learning rule (eRBP) that approximates deep learning in neuromorphic hardware without requiring high-precision or network-wide information.

Findings

eRBP achieves nearly identical accuracy to GPU-based neural networks.

The rule is simple, requiring only one addition and two comparisons per synapse.

Deep representations are learned rapidly and are robust to quantization.

Abstract

An ongoing challenge in neuromorphic computing is to devise general and computationally efficient models of inference and learning which are compatible with the spatial and temporal constraints of the brain. One increasingly popular and successful approach is to take inspiration from inference and learning algorithms used in deep neural networks. However, the workhorse of deep learning, the gradient descent Back Propagation (BP) rule, often relies on the immediate availability of network-wide information stored with high-precision memory, and precise operations that are difficult to realize in neuromorphic hardware. Remarkably, recent work showed that exact backpropagated weights are not essential for learning deep representations. Random BP replaces feedback weights with random ones and encourages the network to adjust its feed-forward weights to learn pseudo-inverses of the (random) feedback weights. Building on these results, we demonstrate an event-driven random BP (eRBP) rule that uses an error-modulated synaptic plasticity for learning deep representations in neuromorphic computing hardware. The rule requires only one addition and two comparisons for each synaptic weight using a two-compartment leaky Integrate & Fire (I&F) neuron, making it very suitable for implementation in digital or mixed-signal neuromorphic hardware. Our results show that using eRBP, deep representations are rapidly learned, achieving nearly identical classification accuracies compared to artificial neural network simulations on GPUs, while being robust to neural and synaptic state quantizations during learning.