TY  - EJOU
AU  - Kakuda, Kazuhiko 
AU  - Enomoto, Tomoyuki 
AU  - Miura, Shinichiro 

TI  - Nonlinear Activation Functions in CNN Based on Fluid Dynamics and Its Applications
T2  - Computer Modeling in Engineering \& Sciences

PY  - 2019
VL  - 118
IS  - 1
SN  - 1526-1506

AB  - The nonlinear activation functions in the deep CNN (Convolutional Neural Network) based on fluid dynamics are presented. We propose two types of activation functions by applying the so-called parametric softsign to the negative region. We use significantly the well-known TensorFlow as the deep learning framework. The CNN architecture consists of three convolutional layers with the max-pooling and one fully-connected softmax layer. The CNN approaches are applied to three benchmark datasets, namely, MNIST, CIFAR-10, and CIFAR-100. Numerical results demonstrate the workability and the validity of the present approach through comparison with other numerical performances.
KW  - Deep learning
KW  -  CNN
KW  -  activation function
KW  -  fluid dynamics
KW  -  MNIST
KW  -  CIFAR-10
KW  -  CIFAR-100

DO  - 10.31614/cmes.2019.04676