4 papers · 1 filter
Stochastic Subnetwork Annealing: A Regularization Technique for Fine Tuning Pruned Subnetworks
Tim Whitaker, Darrell Whitley
Pruning methods have recently grown in popularity as an effective way to reduce the size and computational complexity of deep neural networks. Large numbers of parameters can be re…
Quantum Neuron Selection: Finding High Performing Subnetworks With Quantum Algorithms
Tim Whitaker
Gradient descent methods have long been the de facto standard for training deep neural networks. Millions of training samples are fed into models with billions of parameters, which…
Interpretable Diversity Analysis: Visualizing Feature Representations In Low-Cost Ensembles
Tim Whitaker, Darrell Whitley
Diversity is an important consideration in the construction of robust neural network ensembles. A collection of well trained models will generalize better if they are diverse in th…
Synaptic Stripping: How Pruning Can Bring Dead Neurons Back To Life
Tim Whitaker, Darrell Whitley
Rectified Linear Units (ReLU) are the default choice for activation functions in deep neural networks. While they demonstrate excellent empirical performance, ReLU activations can…