An improved Monte Carlo Tree Search approach to workflow scheduling

Workflow computing has become an essential part of many scientific and engineering fields, while workflow scheduling long been a well-known NP-complete research problem. Major previous works can be classified into two categories: heuristic-based guided random-search-based methods. Monte Carlo Tree Search (MCTS) is recently proposed search methodology with great success in AI on game playing, such as Computer Go. However, researchers found that MCTS also potential application other domains, including combinatorial optimization, task scheduling, planning, so on. In this paper, we present new approach based MCTS, which still rarely explored direction. Several mechanisms are developed for the major steps to improve execution schedules further. Experimental results show our outperforms methods significantly terms makespan.