2 papers
cs.LG2020
Quantal synaptic dilution enhances sparse encoding and dropout regularisation in deep networks
Gardave S Bhumbra
Dropout is a technique that silences the activity of units stochastically while training deep networks to reduce overfitting. Here we introduce Quantal Synaptic Dilution (QSD), a b…
cs.NE2018
Deep learning improved by biological activation functions
Gardave S Bhumbra
`Biologically inspired' activation functions, such as the logistic sigmoid, have been instrumental in the historical advancement of machine learning. However in the field of deep l…