docs: add exploratory statistical significance tutorial

[tomatotomata]

Pull Request Checklist

• Read our Contributing Guide
• Referenced a Github issue: Statistical Significance #731
• Read and reviewed all comments on Statistical Significance #731 and checked cross-referenced work
• Checked that Statistical Significance #731 is not assigned and is not addressed by another pull request
• Left a meaningful approach comment on Statistical Significance #731 and received maintainer confirmation before proceeding
• Forked, cloned, and checked out the newest version of the code
• Created a new branch
• Made the necessary documentation changes
• Installed black (not available in this local environment)
• Installed flake8 (not available in this local environment)
• pytest-cov is available in the local environment
• Ran the notebook JSON and extracted-code checks described below
• Ran the full ./setup.sh dev && ./test.sh suite (the local environment does not have stumpy installed)
• AI code assistance was used to generate 15%+ of this pull request

What does this PR do?

Adds Tutorial_Statistical_Significance.ipynb, a small tutorial based on the approach discussed in #731. It:

• builds a fixed-seed synthetic series with a repeated motif
• compares the observed minimum matrix-profile distance with a null distribution from circular block permutations
• uses a +1 empirical lower-tail correction and shows sensitivity considerations
• documents dependence, multiple comparisons, block-size choices, and the exploratory nature of the result

The notebook is added to the documentation toctree and mirrored through the repository's existing notebooks/ link convention.

Validation

• Notebook JSON parses successfully.
• Extracted code cells compile after removing the IPython magic.
• The control flow was executed with a lightweight brute-force matrix-profile stand-in because stumpy is not installed locally; the example produced a finite empirical p-value and plot construction completed.
• git diff --check passes.

This intentionally stays tutorial-only and does not add a public API. I was thinking of keeping the null construction and caveats visible in the notebook so readers can adapt the method without mistaking it for a universal threshold. Let me know what you think.

Fixes #731

[gitnotebooks]

Found 2 changed notebooks. Review the changes at https://app.gitnotebooks.com/stumpy-dev/stumpy/pull/1172

[github-actions]

❌ This pull request has been automatically closed because the mandatory PR template was removed or not filled out. Please reopen or recreate this PR with the required information.