Skip to content
/ DLTS Public

Deep Learning Assisted Heuristic Tree Search for the Container Pre-marshalling Problem

License

Notifications You must be signed in to change notification settings

ahottung/DLTS

Repository files navigation

DLTS - Deep Learning Assisted Heuristic Tree Search

This implementation was used to conduct the experiments in our paper Deep Learning Assisted Heuristic Tree Search for the Container Pre-marshalling Problem (the preprint of the paper can be found here). Additionally to the code, this repository also contains the validation and test instances (and their optimal solutions) used in the paper.

Dependencies

We evaluated DLTS using Python 3.5 and

  • keras 1.1.0
  • theano 0.8.2
  • h5py 2.7.1

Usage Example

To solve the 5x5_cv1 validation instances using the provided branching and bounding networks use the following command:

python deep_learn.py -t reference_solutions/5x5_cv1_validation -m pre-trained_models/pm_dnn_model_5x5.h5 -s -v pre-trained_models/pm_dnn_value_model_5x5.h5

About

Deep Learning Assisted Heuristic Tree Search for the Container Pre-marshalling Problem

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages