Set Aggregation Network as a Trainable Pooling Layer
This work addresses a fundamental component in neural networks for processing structured data, offering a more flexible and regularizing alternative to standard pooling methods.
The authors tackled the problem of global pooling in deep neural networks by introducing the Set Aggregation Network (SAN) as a trainable pooling layer, which improves classification accuracy and reduces overfitting.
Global pooling, such as max- or sum-pooling, is one of the key ingredients in deep neural networks used for processing images, texts, graphs and other types of structured data. Based on the recent DeepSets architecture proposed by Zaheer et al. (NIPS 2017), we introduce a Set Aggregation Network (SAN) as an alternative global pooling layer. In contrast to typical pooling operators, SAN allows to embed a given set of features to a vector representation of arbitrary size. We show that by adjusting the size of embedding, SAN is capable of preserving the whole information from the input. In experiments, we demonstrate that replacing global pooling layer by SAN leads to the improvement of classification accuracy. Moreover, it is less prone to overfitting and can be used as a regularizer.