ai4co / rl4co

A PyTorch library for all things Reinforcement Learning (RL) for Combinatorial Optimization (CO)
https://rl4.co
MIT License
381 stars 70 forks source link

Fjsp #173

Closed LTluttmann closed 4 months ago

LTluttmann commented 4 months ago

Description

Adding environment for the Flexible Job-Shop Scheduling Problem (FJSP) and the heterogeneous graph neural network (HGNN) described in Song et al. (https://ieeexplore.ieee.org/document/9826438) to solve it. Architecture was improved by adding residual connection and batch normalization. Average makespan on smallest instance type (10 jobs and 5 machines) is ~100 compared to ~106 reported in the paper.

Motivation and Context

closes #168

Types of changes

What types of changes does your code introduce? Remove all that do not apply:

Checklist

Go over all the following points, and put an x in all the boxes that apply. If you are unsure about any of these, don't hesitate to ask. We are here to help!

LTluttmann commented 4 months ago

thanks a ton for the valuable comments @fedebotu & @cbhua! I hope I addressed all issues. Merging now