Ensemble of Convolutional Neural Networks Trained with Different Activation Functions
Ensemble of Convolutional Neural Networks Trained with Different Activation Functions
Gianluca Maguolo,L. Nanni,S. Ghidoni
2019 · DOI: 10.1016/J.ESWA.2020.114048
Expert systems with applications · 69 Citations
TLDR
The goal of this work is to propose an ensemble of Convolutional Neural Networks trained using several different activation functions, and a novel activation function is here proposed for the first time to improve the performance of Convolved Neural Networks in small/medium size biomedical datasets.