A Trial Experiment of Deep Learning For Task Scheduling

Abstract

Task scheduling is one of the methods to minimize the processing time of a program in parallel processing. However, because task scheduling problems belong to strongly NP-hard combinatorial optimization problems [1], it is hard to find an optimal solution within a practical search time. In this research, we performed an experiment for speeding up the traditional search method based on the branch and bound algorithm by conducting scheduling with the use of deep learning.