Predict the Best Graph Partitioning Strategy by Using Machine Learning Technology
Jiayi Shen, Fabrice Huet · 2018
In this paper, we explore applying machine learning techniques to find a best partitioner for a given graph. We use some metrics to describe the graph, and use these metrics as the input and the partitioner ranking of a graph execution algorithm as the label to train a model. Our experiment shows KNN and decision tree are good models for this problem.