Fast dropout training

Preventing feature co-adaptation by encouraging independent contributions from different features often improves classification and regression performance. Dropout training (Hinton et al., 2012) does this by randomly dropping out (zeroing) hidden units and input features during training of neural networks. However, repeatedly sampling a random subset of input features makes training much slower. Based on an examination of the implied objective function of dropout training, we show how to do fast dropout training by sampling from or integrating a Gaussian approximation, instead of doing Monte Carlo optimization of this objective. This approximation, justified by the central limit theorem and empirical evidence, gives an order of magnitude speedup and more stability. We show how to do fast dropout training for classification, regression, and multilayer neural networks. Beyond dropout, our technique is extended to integrate out other types of noise and small image transformations.

[1]  Stephen Tyree,et al.  Learning with Marginalized Corrupted Features , 2013, ICML.

[2]  Michael I. Jordan,et al.  On Discriminative vs. Generative Classifiers: A comparison of logistic regression and naive Bayes , 2001, NIPS.

[3]  David J. C. MacKay,et al.  The Evidence Framework Applied to Classification Networks , 1992, Neural Computation.

[4]  Christopher M. Bishop,et al.  Current address: Microsoft Research, , 2022 .

[5]  Yoshua Bengio,et al.  Maxout Networks , 2013, ICML.

[6]  Yann LeCun,et al.  Transformation Invariance in Pattern Recognition-Tangent Distance and Tangent Propagation , 1996, Neural Networks: Tricks of the Trade.

[7]  Andrew McCallum,et al.  Reducing Weight Undertraining in Structured Discriminative Learning , 2006, NAACL.

[8]  Andrew M. Ross.  Computing Bounds on the Expected Maximum of Correlated Normal Variables , 2010 .

[9]  Nitish Srivastava,et al.  Improving neural networks by preventing co-adaptation of feature detectors , 2012, ArXiv.

[10]  Christopher Potts,et al.  Learning Word Vectors for Sentiment Analysis , 2011, ACL.

[11]  Christopher D. Manning,et al.  Baselines and Bigrams: Simple, Good Sentiment and Topic Classification , 2012, ACL.

[12]  E. Lehmann.  Elements of large-sample theory , 1998 .

[13]  Kiyotoshi Matsuoka,et al.  Noise injection into inputs in back-propagation learning , 1992, IEEE Trans. Syst. Man Cybern..

[14]  Kentaro Inui,et al.  Dependency Tree-based Sentiment Classification using CRFs with Hidden Variables , 2010, NAACL.

[15]  Jeffrey Pennington,et al.  Semi-Supervised Recursive Autoencoders for Predicting Sentiment Distributions , 2011, EMNLP.

[16]  Trevor Cohn,et al.  Logarithmic Opinion Pools for Conditional Random Fields , 2005, ACL.