Event-based backpropagation can compute exact gradients for spiking neural networks
Open Access
- 18 June 2021
- journal article
- research article
- Published by Springer Science and Business Media LLC in Scientific Reports
- Vol. 11 (1), 1-17
- https://doi.org/10.1038/s41598-021-91786-z
Abstract
Spiking neural networks combine analog computation with event-based communication using discrete spikes. While the impressive advances of deep learning are enabled by training non-spiking artificial neural networks using the backpropagation algorithm, applying this algorithm to spiking networks was previously hindered by the existence of discrete spike events and discontinuities. For the first time, this work derives the backpropagation algorithm for a continuous-time spiking neural network and a general loss function by applying the adjoint method together with the proper partial derivative jumps, allowing for backpropagation through discrete spike events without approximations. This algorithm, EventProp, backpropagates errors at spike times in order to compute the exact gradient in an event-based, temporally and spatially sparse fashion. We use gradients computed via EventProp to train networks on the Yin-Yang and MNIST datasets using either a spike time or voltage based loss function and report competitive performance. Our work supports the rigorous study of gradient-based learning algorithms in spiking neural networks and provides insights toward their implementation in novel brain-inspired hardware.Other Versions
This publication has 48 references indexed in Scilit:
- The Chronotron: A Neuron That Learns to Fire Temporally Precise Spike PatternsPLOS ONE, 2012
- The tempotron: a neuron that learns spike timing–based decisionsNature Neuroscience, 2006
- A gradient descent rule for spiking neurons emitting multiple spikesInformation Processing Letters, 2005
- Modeling, simulation, sensitivity analysis, and optimization of hybrid systemsACM Transactions on Modeling and Computer Simulation, 2002
- Learning Methods of Recurrent Spiking Neural NetworksTransactions of the Institute of Systems, Control and Information Engineers, 2000
- Parametric sensitivity functions for hybrid discrete/continuous systemsApplied Numerical Mathematics, 1999
- Dynamic Optimization in a Discontinuous WorldIndustrial & Engineering Chemistry Research, 1998
- Gradient-based learning applied to document recognitionProceedings of the IEEE, 1998
- Networks of spiking neurons: The third generation of neural network modelsNeural Networks, 1997
- Note on the Derivatives with Respect to a Parameter of the Solutions of a System of Differential EquationsAnnals of Mathematics, 1919