Created by Team Qunonians for the D-Wave Challenge @ iQuHack 2025
Authors: Atharv Chowdhary, Rohan Kher, Leopold Li, Simon Nirenberg, Jerry Sun
We take on D-Wave's harder challenge of implementing a solution to an advanced version of the Quadratic Assignment Problem (QAP). We introduce a method for combining matrices and penalizing select cells, which allows us to solve QAP in time dependent systems, evaluate costs involved in transitioning between states, and distinguish between various kinds of actors involved in the problem formulation. We apply this method on D-Wave's Quantum Annealing systems.
We offer two key deliverables as part of our submission: a Jupyter notebook which walks the user through our methodology, and a Python library which serves as a helpful abstraction for solving similar QAP problems with D-Wave's Quantum Computers.
Our first key deliverable for this project is Jupyter notebook qap_theory.ipynb, located in this repository at this link. This notebook goes into extensive detail as to the theory behind our algorithm, and our methodology for its implementation. It covers the key details behind the math, physics, and computer science that makes our algorithm function.
Our second key deliverable for this project is the library ocean_qap, which primarily builds off code from the qap.py file in this repository. A link to the complete repository for ocean_qap can be found here, and the library can be installed via this link or using the command pip install ocean-qap.
ocean_qap provides a helpful abstraction via the QAP class, which conveniently stores and handles all the required information required for solving Quadratic Assignment Problems of the nature described by the challenge. Users can easily initialize the class, obtain optimizations, and evolve the system to obtain new optimizations based on new parameters and transition states. Various customization and visualization options are also available. A brief tutorial can be found in the ocean_qap.ipynb notebook located in this repository, at this link.