Adaptive hybrid activation function for deep neural networks

Autor: Yevgeniy Bodyanskiy, Serhii Kostiuk
Jazyk: ukrajinština
Rok vydání: 2022
Předmět:
Zdroj: Sistemnì Doslìdženâ ta Informacìjnì Tehnologìï, Iss 1 (2022)
Druh dokumentu: article
ISSN: 2308-8893
1681-6048
DOI: 10.20535/SRIT.2308-8893.2022.1.07
Popis: The adaptive hybrid activation function (AHAF) is proposed that combines the properties of the rectifier units and the squashing functions. The proposed function can be used as a drop-in replacement for ReLU, SiL and Swish activations for deep neural networks and can evolve to one of such functions during the training. The effectiveness of the function was evaluated on the image classification task using the Fashion-MNIST and CIFAR-10 datasets. The evaluation shows that the neural networks with AHAF activations achieve better classification accuracy comparing to their base implementations that use ReLU and SiL. A double-stage parameter tuning process for training the neural networks with AHAF is proposed. The proposed approach is sufficiently simple from the implementation standpoint and provides high performance for the neural network training process.
Databáze: Directory of Open Access Journals