When Expressivity Meets Trainability: Fewer than $n$ Neurons Can Work.

Jiawei Zhang Yushun Zhang Mingyi Hong Ruoyu Sun Zhi-Quan Luo

Published in: NeurIPS (2021)

Keyphrases

neural network
hidden layer
spiking neurons
visual cortex
receptive fields
neuron model
neural models
data sets
associative memory
feed forward
activation function
learning rules
multiscale
spiking neural networks
case study
decision trees
single neuron
machine learning
synaptic plasticity