Adds MAPE function to Loss functions

[morgen-code]

Describe your change:

Fixes #13311
Adds the algorithm to calculate mean_absolute_percentage_error in loss_functions.py file in machine_learning directory.

• Add an algorithm?
• Fix a bug or typo in an existing algorithm?
• Add or change doctests? -- Note: Please avoid changing both code and tests in a single pull request.
• Documentation change?

Checklist:

• I have read CONTRIBUTING.md.
• This pull request is all my own work -- I have not plagiarized.
• I know that pull requests will not be merged if they fail the automated tests.
• This PR only changes one algorithm file. To ease review, please open separate PRs for separate algorithms.
• All new Python files are placed inside an existing directory.
• All filenames are in all lowercase characters with no spaces or dashes.
• All functions and variable names follow Python naming conventions.
• All function parameters and return values are annotated with Python type hints.
• All functions have doctests that pass the automated testing.
• All new algorithms include at least one URL that points to Wikipedia or another similar explanation.
• If this pull request resolves one or more open issues then the description above includes the issue number(s) with a closing keyword: "Fixes #ISSUE-NUMBER".

[priya-sundaram-dev]

Thanks for the PR! The math is right, but CI is red (ruff / pre-commit / build) and there are a few copy-paste issues to fix before this can go in:

1. Doctests call the wrong function. They invoke symmetric_mean_absolute_percentage_error(...), but this PR defines mean_absolute_percentage_error. So the doctests are actually testing SMAPE, not MAPE. Change the calls to mean_absolute_percentage_error. (The expected value 0.058333… is in fact correct for MAPE on that data — 0.1, 0.05, 0.0333, 0.05 averaged — so just fixing the function name makes it pass.)
2. Docstring formula/name. It reads SMAPE = …; it should say MAPE = (1/n) * Σ( |y_true - y_pred| / |y_true| ).
3. Indentation. The docstring body is over-indented with mixed leading whitespace, which is part of the lint failure. Align everything to 4 spaces under the opening """.
4. Line length. The >>> float(mean_absolute_percentage_error(true_values, predicted_values)) line will likely exceed the 88-char limit — shorten the local variable names (e.g. true/pred) so it fits.

Once those are fixed the doctests and ruff should pass. Also heads-up: #13355 (SMAPE) touches the same block of loss_functions.py, so a rebase may be needed depending on merge order.

[cclauss]

@morgen-code @priya-sundaram-dev Can either of you manage this rebase?

[priya-sundaram-dev]

Before anyone spends time on the rebase — heads up that this one looks superseded. machine_learning/loss_functions.py on master already defines mean_absolute_percentage_error (currently around line 439, added in an earlier PR), with the same (y_true, y_pred, epsilon=1e-15) signature and passing doctests. That's also the source of the conflict here.

So rather than rebasing, I think #13357 can be closed as a duplicate. Two small notes for the record, in case anything from here is worth folding into the existing function:

• The doctests in this PR call symmetric_mean_absolute_percentage_error(...) rather than mean_absolute_percentage_error(...), and the formula comment says SMAPE = ... — looks like a copy-paste from the SMAPE PR, so they wouldn't pass as written.
• The existing master version already handles the divide-by-zero guard via epsilon.

Thanks @morgen-code for the contribution and the interest in the loss-function module! Happy to help if you'd like to pick up a different addition. @cclauss let me know if you'd prefer I do anything else here.