Stochastic Optimization of Plain Convolutional Neural Networks with Simple methods
This work addresses regularization for CNNs in visual classification, but it is incremental as it combines existing techniques without introducing new methods.
The authors tackled overfitting in plain convolutional neural networks by combining data augmentation, dropout, and a customized early stopping function, achieving state-of-the-art results on MNIST, SVHN, and STL10 datasets and high accuracy on CIFAR10 and CIFAR100.
Convolutional neural networks have been achieving the best possible accuracies in many visual pattern classification problems. However, due to the model capacity required to capture such representations, they are often oversensitive to overfitting and therefore require proper regularization to generalize well. In this paper, we present a combination of regularization techniques which work together to get better performance, we built plain CNNs, and then we used data augmentation, dropout and customized early stopping function, we tested and evaluated these techniques by applying models on five famous datasets, MNIST, CIFAR10, CIFAR100, SVHN, STL10, and we achieved three state-of-the-art-of (MNIST, SVHN, STL10) and very high-Accuracy on the other two datasets.