shapiq.approximator.SVARM

class shapiq.approximator.SVARM(n, index='SV', *, random_state=None, pairing_trick=False, sampling_weights=None, **kwargs)

Bases: MonteCarlo[Literal[‘SV’, ‘BV’]]

The SVARM approximator for estimating the Shapley value (SV).

SVARM is a MonteCarlo approximation algorithm that estimates the Shapley value. For details about the algorithm see the original paper by Kolpaczki et al. (2024) Kolpaczki et al.[1].

References

[1]

Patrick Kolpaczki, Viktor Bengs, Maximilian Muschalik, and Eyke Hüllermeier. Approximating the shapley value without marginal contributions. In Thirty-Eighth AAAI Conference on Artificial Intelligence, (AAAI 2024), 13246–13255. AAAI Press, 2024. doi:10.1609/AAAI.V38I12.29225.

Initialize the SVARM approximator.

Parameters:

• n (int) – The number of players.

• index (Literal['SV', 'BV']) – The interaction index to be used. Choose from ['SV', 'BV']. Defaults to 'SV'.

• random_state (int | None) – The random state of the estimator. Defaults to None.

• pairing_trick (bool) – If True, the pairing trick is applied to the sampling procedure. Defaults to False.

• sampling_weights (ndarray[tuple[Any, ...], dtype[floating]] | None) – An optional array of weights for the sampling procedure. The weights must be of shape (n + 1,) and are used to determine the probability of sampling a coalition of a certain size. Defaults to None.

• **kwargs (Any) – Additional keyword arguments (not used only for compatibility).

valid_indices: tuple[Literal['SV', 'BV'], ...] = ('SV', 'BV')

The valid indices for the SVARM approximator.