Add demo: Qubit-Efficient Encoding Techniques for Solving QUBO Problems

[supreethmv]

Title: Add demo: Qubit-Efficient Encoding Techniques for Solving QUBO Problems

Summary:

This demo showcases variational quantum algorithms (VQAs) for solving combinatorial optimization problems using qubit-efficient encoding strategies. The demonstration is centered around an image segmentation task, where the segmentation problem is formulated as a minimum cut, then cast as a QUBO and solved using logarithmic-qubit methods.

The notebook walks through three methods:

• Parametric Gate Encoding (PGE)
• Ancilla Basis Encoding (ABE)
• Adaptive Cost Encoding (ACE, a new technique introduced in our paper)

Compared to QAOA, these techniques scale much better in terms of quantum resources and are well-suited for current and near-term quantum devices.

Relevant references:

• Venkatesh, S. M., et al., "Qubit-Efficient Variational Quantum Algorithms for Image Segmentation", IEEE QCE 2024. DOI: 10.1109/QCE60285.2024.00059

Possible Drawbacks:

None expected. The code is tested, PEP8-compliant, and figures are formatted using the required .. figure:: syntax. Metadata and preview image are included.

Related GitHub Issues:

N/A

---

Demo metadata

• GOALS — Showcase a scalable PennyLane-based implementation of resource-efficient VQAs, useful for real-world QUBO problems beyond toy examples. Also demonstrates PL as a platform to prototype cutting-edge QML research.

• AUDIENCE — Quantum ML researchers, optimization specialists, educators teaching VQAs, and developers looking for practical NISQ-friendly encodings.

• KEYWORDS — qubit-efficient, VQA, QUBO, image segmentation, minimum cut, ACE, PGE, ABE, QAOA alternative, NISQ

• Documentation type:

• Demo

[github-actions]

Thank you for opening this pull request.

You can find the built site at this link.

Deployment Info:

• Pull Request ID: 1343
• Deployment SHA: 5e151bbc5d42ae82b65360c2bf2bbdba3c35f8f7
  (The Deployment SHA refers to the latest commit hash the docs were built from)

Note: It may take several minutes for updates to this pull request to be reflected on the deployed site.

[CatalinaAlbornoz]

Hi @supreethmv, thank you for submitting this demo!

At the moment we've got several demos lined up in our roadmap for publication, so it might be best if we add your demo in our community demos page for now. This means that you can host the demo on your own repository, and we can add it to the site faster. Would this be ok with you? We can still keep this PR open so that we can add it in the main demos page once we have capacity.

If you're ok with starting with the community demo we'd need the following information so that we can add it to the website and do a marketing post for it.

Name
Your full name (or username).

Affiliation (optional)
Your affiliation, if applicable; e.g. University, research institute, company.

Socials (optional)
Your Twitter and LinkedIn username, if interested; helps us advertise your demo while linking directly back to you. If you're on the Xanadu Slack we can tag you there too.

Image (optional)
Suggested image to use when advertising your demo on Twitter; can be provided via hyperlink or by copy/pasting directly in GitHub.

Relevant link
Add a link to your demo (as a GitHub repository, Jupyter notebook, Python script, etc.).

Let us know if you have any questions!

[supreethmv]

Hi @supreethmv, thank you for submitting this demo!

At the moment we've got several demos lined up in our roadmap for publication, so it might be best if we add your demo in our community demos page for now. This means that you can host the demo on your own repository, and we can add it to the site faster. Would this be ok with you? We can still keep this PR open so that we can add it in the main demos page once we have capacity.

Hi @CatalinaAlbornoz, thank you for the suggestion. I have now also opened the issue #1345 for submitting this demo as a community demo,

[cognigami]

This was submitted as a Community Demo, and is now live.

[supreethmv]

This was submitted as a Community Demo, and is now live.

Hi @cognigami, thank you for publishing it as a Community Demo, I appreciate the support!
That said, I had spent considerable time making sure the demo conforms to the full PennyLane demo guidelines (including formatting, metadata, ReST figures, etc.), and I was under the impression, based on my discussion above with @CatalinaAlbornoz that the original PR would remain open so it could be merged into the main demos page once there was capacity.

I’d be grateful if the original PR could be reopened and considered for eventual inclusion as a main demo when the queue allows.

[josh146]

Hi @supreethmv of course, I've re-opened the PR!

[supreethmv]

Thank you @josh146 for the quick response!

[cognigami]

This was posted as a community demo.

[CatalinaAlbornoz]

Hi @cognigami, while the demo was posted as a community demo we also want to keep this PR open. Both types of demos have a different style and it might be valuable to have this one too when we have capacity in the team to review and add to the proper category on the main demos page. So let's reopen this PR and keep it open unless there's an additional motivation to close it.