Graph-based Reinforcement Learning with Permutation-Invariant Resource Encoding for Resource-Constrained Unrelated Parallel Machine Scheduling

Authors: Jimin Park, Yongjae Kim, Minju Jang, Min Geon Kim, Yeo Reum Kim, Soyoung Lee, Jihyun Jo, Iksoo Song, Hyunsoon Park, and Hyun-Jung Kim Status: Working paper Collaboration: LG Electronics

Overview

✍️ Add a short summary of the problem, the approach (graph-based RL with a permutation-invariant resource encoding), and the main results here.

My Role

I am the first author of this work and led the research — from problem formulation and method design to implementation, experiments, and writing — in collaboration with LG Electronics.

✍️ Add specifics (e.g. the resource-encoding design, the RL formulation, and what was validated on the production data).