Learning 2opt Drl Save
Learning 2-opt Heuristics for the TSP via Deep Reinforcement Learning
Project README
Learning 2-opt Heuristics for the TSP via Deep Reinforcement Learning 
Implementation of the Policy Gradient algorithm for learning 2-opt improvement heuristics, following: http://proceedings.mlr.press/v129/costa20a/costa20a.pdf
https://arxiv.org/abs/2004.01608
Dependencies:
- Python 3.6.4
- Torch
- Numpy
- Matplotlib
- Apex
- tqdm
- pyconcorde
How to train it?
To train the model you can run:
For TSP instances with 20 nodes:
python PGTSP20.py
For TSP instances with 50/100 nodes (default 50 nodes):
python PGTSP50_100.py
How to test it?
To use the learned polcies reported in the paper you can run:
python TestLearnedAgent.py --load_path best_policy/policy-TSP20-epoch-189.pt --n_points 20 --test_size 1 --render
where load_path can be replaced with one of the policies in /best_policy.
Results
Learned policy on a TSP with 50 nodes:
Open Source Agenda is not affiliated with "Learning 2opt Drl" Project. README Source: paulorocosta/learning-2opt-drl
Stars
41
Open Issues
1
Last Commit
3 years ago
Repository
Tags
2 Opt
Actor Critic
Combinatorial Optimization
Deep Learning
Deep Neural Networks
Graph
Graph Neural Networks
Lstm
Lstm Neural Networks
Policy Gradient
Pytorch
Reinforce
Reinforcement Learning
Reinforcement Learning Agent
Reinforcement Learning Algorithms
Reinforcement Learning Environments
Traveling Salesman
Travelling Salesman Problem
Tsp
Tsp Instances
Open Source Agenda Badge
