[2E4-OS-9] 顔文字の科学

[2E4-OS-9-02] A Multi-class Classification based on the Original Form of Kaomoji using Neural Network

〇Noriyuki Okumura¹, Rei Okumura² (1. Otemae University, 2. National Institute of Technology, Akashi College)

Keywords:Kaomoji, Neural Network

In this paper, we propose a multi-class classification method for Kaomoji using feed-forward neural network.
Neural network has some units in each layer, but a suitable number of units is not precise.
This research investigated the relation between the number of units and the accuracy of the multi-class classification method. As a result, we found out the suitable number of units in a hidden layer is 6,500 based on a rule of thumb in our research.