added multi armed bandit problem with three strategies to solve it - #12668

🔗 Relevant Links

Repository:

• Contributing guidelines
• Project Euler solution guidelines

Python:

• Formatted string literals (f-strings)
• Type hints
• doctest
• unittest
• pytest

algorithms-keeper actions can be triggered by commenting on this PR:

• @algorithms-keeper review to trigger the checks for only added pull request files
• @algorithms-keeper review-all to trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.

NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.

🔗 Relevant Links

Repository:

• Contributing guidelines
• Project Euler solution guidelines

Python:

• Formatted string literals (f-strings)
• Type hints
• doctest
• unittest
• pytest

algorithms-keeper actions can be triggered by commenting on this PR:

• @algorithms-keeper review to trigger the checks for only added pull request files
• @algorithms-keeper review-all to trigger the checks for all the pull request files, including the modified files. As we cannot post review comments on lines not part of the diff, this command will post all the messages in one comment.

NOTE: Commands are in beta and so this feature is restricted only to a member or owner of the organization.

Adds a NumPy-based multi-armed bandit implementation with several strategies and a simulation visualization.

Changes:

• Adds bandit environment and selection strategies.
• Adds doctests and cumulative-reward plotting.
• Updates the generated directory index.

Suppressed comments (6)

machine_learning/mab.py:109

• This constructor has no doctest, although the repository guideline requires every function to have at least one passing doctest. Add a small initialization example here so this new public API is covered directly.

    def __init__(self, epsilon: float, num_arms: int) -> None:
        """
        Initialize the Epsilon-Greedy strategy.

        Args:
            epsilon: The probability of exploring new arms.
            num_arms: The number of arms.
        """

machine_learning/mab.py:169

• This constructor has no doctest, although the repository guideline requires every function to have at least one passing doctest. Add a small initialization example here so this new public API is covered directly.

    def __init__(self, num_arms: int) -> None:
        """
        Initialize the UCB strategy.

        Args:
            num_arms: The number of arms.
        """

machine_learning/mab.py:228

• This constructor has no doctest, although the repository guideline requires every function to have at least one passing doctest. Add a small initialization example here so this new public API is covered directly.

    def __init__(self, num_arms: int) -> None:
        """
        Initialize the Thompson Sampling strategy.

        Args:
            num_arms: The number of arms.
        """

machine_learning/mab.py:287

• This constructor has no doctest, although the repository guideline requires every function to have at least one passing doctest. Add a small initialization example here so this new public API is covered directly.

    def __init__(self, num_arms: int) -> None:
        """
        Initialize the Random strategy.

        Args:
            num_arms: The number of arms.
        """

machine_learning/mab.py:334

• This constructor has no doctest, although the repository guideline requires every function to have at least one passing doctest. Add a small initialization example here so this new public API is covered directly.

    def __init__(self, num_arms: int) -> None:
        """
        Initialize the Greedy strategy.

        Args:
            num_arms: The number of arms.
        """

machine_learning/mab.py:107

• The parameter description says epsilon explores “new arms,” but the implementation samples any arm, including one already tried. Document this as choosing a random arm so callers do not infer an unimplemented untried-arm guarantee.

            epsilon: The probability of exploring new arms.

• Files reviewed: 2/2 changed files
• Comments generated: 7
• Review effort level: Lite