Ghost Units Yield Biologically Plausible Backprop in Deep Neural Networks.
Thomas MesnardGaëtan VignoudJoão SacramentoWalter SennYoshua BengioPublished in: CoRR (2019)
Keyphrases
- biologically plausible
- feed forward
- neural network
- biologically inspired
- network architecture
- back propagation
- synaptic weights
- visual cortex
- spiking neurons
- spiking neural networks
- artificial neural networks
- neural nets
- recurrent neural networks
- hidden layer
- neural network model
- pattern recognition
- multilayer perceptron
- learning rules
- multi modal
- biological systems
- fuzzy logic
- fault diagnosis
- receptive fields
- decision making
- computer vision
- genetic algorithm