Open AccessPerformance of Different Graph Neural Networks on Graph Classification
Author(s) -
Zikang Zhan
Publication year - 2020
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1607/1/012091
Subject(s) - computer science , convolutional neural network , artificial intelligence , multilayer perceptron , recurrent neural network , artificial neural network , graph , time delay neural network , deep learning , pattern recognition (psychology) , machine learning , theoretical computer science
In recent years, machine learning has become a common approach for solving problems of artificial intelligence, which is developing at a high speed. Convolutional Neural Network (CNN) is popular on solving image problems for its believable accuracy. However, it does not mean that other neural networks are useless on these problems. In this paper, performance of Convolutional Neural Network (CNN), Multilayer Perceptron (MLP) and Recurrent Neural Network (RNN) in the same database are presented. Performance of different types of networks are compared, and a conclusion is given that recurrent neural network might be able to classify pure black and white images at high accuracy, and that convolutional neural network performs better than multilayer perceptron on graph classification problems.