Max-Pooling Dropout for Regularization of Convolutional Neural Networks

TL;DR

This paper introduces probabilistic weighted pooling as a regularization technique for convolutional neural networks, demonstrating its effectiveness over traditional max-pooling and stochastic pooling methods through empirical validation.

Contribution

It reveals the equivalence of max-pooling dropout to multinomial sampling and proposes probabilistic weighted pooling as a novel, superior regularization method for CNNs.

Findings

Probabilistic weighted pooling outperforms max-pooling in regularization.

Max-pooling dropout is equivalent to multinomial sampling.

Empirical results show improved accuracy with probabilistic weighted pooling.

Abstract

Recently, dropout has seen increasing use in deep learning. For deep convolutional neural networks, dropout is known to work well in fully-connected layers. However, its effect in pooling layers is still not clear. This paper demonstrates that max-pooling dropout is equivalent to randomly picking activation based on a multinomial distribution at training time. In light of this insight, we advocate employing our proposed probabilistic weighted pooling, instead of commonly used max-pooling, to act as model averaging at test time. Empirical evidence validates the superiority of probabilistic weighted pooling. We also compare max-pooling dropout and stochastic pooling, both of which introduce stochasticity based on multinomial distributions at pooling stage.