AnugyaSahu / Combinatorial-problems-with-Transformers

Solving complex real-world COPs with limited data / information and deep learning
Apache License 2.0
0 stars 0 forks source link

Look for real-world COPs #3

Closed riridi closed 1 year ago

riridi commented 1 year ago

As said, I am especially interested in real-world COPs (without trivial cost function or heuristic). The only dataset I currently know of is the PROCAT dataset: https://github.com/mateuszjurewicz/procat
Could you look for other datasets like that or if there even datasets like that?
One idea I had was looking for real-world vehicle routing datasets or image jigsaw puzzels. NLP problems are also related to COPs. But always think about the properties of the problem (input, output, complexity, solutions, features).

AnugyaSahu commented 1 year ago

Real world VRP datasets -

http://people.brunel.ac.uk/~mastjjb/jeb/orlib/areainfo.html http://www.raucq.be/ (Complex, different tables, preprocessing required, high number of features , solutions/output sizes are complex too relating to different vehicles or depots)

Papers/Datasets on image jigsaw puzzle-

https://www.kaggle.com/datasets/serhiibiruk/jigsaw-puzzle https://arxiv.org/pdf/2101.07555.pdf https://arxiv.org/pdf/2207.11971.pdf https://deepai.org/publication/deepzzle-solving-visual-jigsaw-puzzles-with-deep-learning-andshortest-path-optimization (input and output are images, solutions already available, features are varied according to image size and divisions into the puzzle and complexity also is proportional to features)