Enhancing the Regularization Effect of Weight Pruning in Artificial Neural Networks
This addresses computational and memory costs for mobile or embedded devices by enhancing regularization in weight pruning.
The paper tackled the problem of improving generalization in pruned artificial neural networks by targeting large weights instead of small ones, resulting in higher image classification accuracy on CIFAR-10 than dropout and an 85% reduction in parameter count.
Artificial neural networks (ANNs) may not be worth their computational/memory costs when used in mobile phones or embedded devices. Parameter-pruning algorithms combat these costs, with some algorithms capable of removing over 90% of an ANN's weights without harming the ANN's performance. Removing weights from an ANN is a form of regularization, but existing pruning algorithms do not significantly improve generalization error. We show that pruning ANNs can improve generalization if pruning targets large weights instead of small weights. Applying our pruning algorithm to an ANN leads to a higher image classification accuracy on CIFAR-10 data than applying the popular regularizer dropout. The pruning couples this higher accuracy with an 85% reduction of the ANN's parameter count.