Improving neural networks by preventing co-adaptation of feature detectors

TL;DR

This paper introduces dropout, a regularization technique that randomly omits feature detectors during training to prevent co-adaptation, significantly improving neural network performance on various tasks.

Contribution

The paper presents dropout as a novel regularization method that reduces overfitting by preventing co-adaptation of neurons in neural networks.

Findings

Dropout reduces overfitting on small datasets.

Significant performance improvements on speech and object recognition benchmarks.

Sets new state-of-the-art results with dropout implementation.

Abstract

When a large feedforward neural network is trained on a small training set, it typically performs poorly on held-out test data. This "overfitting" is greatly reduced by randomly omitting half of the feature detectors on each training case. This prevents complex co-adaptations in which a feature detector is only helpful in the context of several other specific feature detectors. Instead, each neuron learns to detect a feature that is generally helpful for producing the correct answer given the combinatorially large variety of internal contexts in which it must operate. Random "dropout" gives big improvements on many benchmark tasks and sets new records for speech and object recognition.