Discovering Parametric Activation Functions
This addresses the challenge of selecting effective activation functions for deep learning practitioners, offering an automated optimization step for new tasks, though it is incremental as it builds on existing activation function research.
The paper tackles the problem of inconsistent benefits from novel activation functions in deep learning by proposing a technique to automatically customize activation functions, resulting in reliable performance improvements, such as consistent accuracy gains over ReLU and other functions on CIFAR-10 and CIFAR-100 datasets.
Recent studies have shown that the choice of activation function can significantly affect the performance of deep learning networks. However, the benefits of novel activation functions have been inconsistent and task dependent, and therefore the rectified linear unit (ReLU) is still the most commonly used. This paper proposes a technique for customizing activation functions automatically, resulting in reliable improvements in performance. Evolutionary search is used to discover the general form of the function, and gradient descent to optimize its parameters for different parts of the network and over the learning process. Experiments with four different neural network architectures on the CIFAR-10 and CIFAR-100 image classification datasets show that this approach is effective. It discovers both general activation functions and specialized functions for different architectures, consistently improving accuracy over ReLU and other activation functions by significant margins. The approach can therefore be used as an automated optimization step in applying deep learning to new tasks.