Graph neural network for graph partitioning

Abstract

Graph partitioning, also known as clustering, has always been an important subject of data analysis and is currently applied in different fields in reality. For example, in business, cluster analysis is used to find different customer groups and distinguish the characteristics of different customer groups through purchase patterns. In biology, cluster analysis is used to classify animal and plant genes to gain insights into biological structure. The definition of clustering is grouping similar objects into the same group. This project aims to implement different clustering methods, one is based on graph neural network (GCN), and the others are based on classic clustering algorithms (K-mean, Hierarchical, etc.). In this project, the sample data will be represented in graph form and then perform clustering by Python. Then make comparisons for all clustering methods to evaluate the results based on standards such as execution time and accuracy to know the performance of different clustering methods.