Event-Driven Random Back-Propagation: Enabling Neuromorphic Deep Learning Machines
TON DUC THANG University
Event-Driven Random Back-Propagation: Enabling Neuromorphic Deep Learning Machines
Author:
Neftci, Emre O
;
Augustine, Charles
;
Paul, Somnath
;
Detorakis, Georgios
Subjects:
Artificial intelligence
;
Back propagation
;
backpropagation algorithm
;
Brain research
;
Computers
;
Deep learning
;
Distance learning
;
embedded cognition
;
Feedback
;
feedback alignment
;
Firing pattern
;
International conferences
;
Learning algorithms
;
Machine learning
;
Memory
;
Neural networks
;
Neurons
;
Neuroscience
;
Neurosciences
;
Signal processing
;
spiking neural networks
;
stochastic processes
;
Synaptic plasticity
;
Synaptic strength
Is Part Of:
Frontiers in neuroscience, 2017-06, Vol.11, p.324-324
Description:
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 Gradient Back Propagation (BP) rule, often relies on the immediate availability of network-wide information stored with high-precision memory during learning, and precise operations that are difficult to realize in neuromorphic hardware. Remarkably, recent work showed that exact backpropagated gradients are not essential for learning deep representations. Building on these results, we demonstrate an event-driven random BP (eRBP) rule that uses an error-modulated synaptic plasticity for learning deep representations. Using a two-compartment Leaky Integrate & Fire (I&F) neuron, the rule requires only one addition and two comparisons for each synaptic weight, 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 classification accuracies on permutation invariant datasets comparable to those obtained in artificial neural network simulations on GPUs, while being robust to neural and synaptic state quantizations during learning.
Publisher:
Switzerland: Frontiers Research Foundation
Language:
English
Identifier:
ISSN: 1662-4548
ISSN: 1662-453X
EISSN: 1662-453X
DOI: 10.3389/fnins.2017.00324
PMID: 28680387
Source:
Freely Accessible Journals
AUTh Library subscriptions: ProQuest Central
PubMed Central
DOAJ Directory of Open Access Journals