Bayesian Convolutional Neural Networks with Bernoulli Approximate Variational Inference
This addresses the challenge of using CNNs with limited labeled data, which is common in applications where data collection is difficult, though it is incremental as it builds on existing Bayesian and dropout methods.
The paper tackles the problem of convolutional neural networks overfitting on small datasets by introducing a Bayesian CNN with Bernoulli approximate variational inference, showing a considerable improvement in classification accuracy and improving on state-of-the-art results for CIFAR-10.
Convolutional neural networks (CNNs) work well on large datasets. But labelled data is hard to collect, and in some applications larger amounts of data are not available. The problem then is how to use CNNs with small data -- as CNNs overfit quickly. We present an efficient Bayesian CNN, offering better robustness to over-fitting on small data than traditional approaches. This is by placing a probability distribution over the CNN's kernels. We approximate our model's intractable posterior with Bernoulli variational distributions, requiring no additional model parameters. On the theoretical side, we cast dropout network training as approximate inference in Bayesian neural networks. This allows us to implement our model using existing tools in deep learning with no increase in time complexity, while highlighting a negative result in the field. We show a considerable improvement in classification accuracy compared to standard techniques and improve on published state-of-the-art results for CIFAR-10.