Doubt about number of occurrences in the DWAVE output

We are working on solving the Traveling Salesman Problem using the DWAVE Advantadge_system4.1. We have using the QUBO formulation built in https://www.frontiersin.org/articles/10.3389/fict.2019.00013/full.

We have a problem which basically consist on having a very sparse distribution of results (despite the DWAVE computer reaches the optimum result). In the following example, we can see how the computer has reached the optimum value just in two occasions, while the rest of the runs has reached worse results in a very sparse distribution.

Where can we found information about this issue and how to solve it? Is there any documentation or paper which we can read?

Many thanks in advance.

0

Comments

1 comment
  • Eneko,

    While it is hard to say exactly what is happening without seeing how you implemented the algorithm or many other details, one possibility is that the weight of your objective function with respect to the penalty/constraint functions may need some tuning.

    Another possibility includes trying various chain strengths, to find the balance between minimizing chain breaks and at the other end, having the qubit values freeze out too soon and not interact with the others.

    Also, some heavily constrained problems may return only a few maximally optimal solutions.

    Depending on how you coded your solution, the above items may not be an issue, since some of the Ocean functions tune the objective weights and chain strengths for you.

    Again, these are just guesses as to what might be causing the distribution you are seeing.

    Good luck on your problem.

    Ed

     

    0
    Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post