KernelSHAP

Regression-based Shapley value approximation using KernelSHAP Lundberg and Lee[1].

from __future__ import annotations

import numpy as np

import shapiq

N_PLAYERS = 8
BUDGET = 200
feature_names = [f"x{i}" for i in range(N_PLAYERS)]

weights = np.array([0.4, 0.3, 0.2, 0.1, 0.05, -0.1, -0.2, -0.3])


def game_fun(coalitions: np.ndarray) -> np.ndarray:
    coalitions = np.atleast_2d(coalitions)
    return (coalitions @ weights) + 0.5 * coalitions[:, 0] * coalitions[:, 1]

Approximate Shapley values

approximator = shapiq.KernelSHAP(n=N_PLAYERS, random_state=42)
iv = approximator.approximate(BUDGET, game_fun)
print(iv)

InteractionValues(
    index=SV, max_order=1, min_order=0, estimated=True, estimation_budget=200,
    n_players=8, baseline_value=0.0,
    Top 10 interactions:
        (0,): 0.6455994941730212
        (1,): 0.557218517264531
        (2,): 0.18232714504237765
        (3,): 0.09776947007519904
        (4,): 0.036511144943958726
        (): 0.0
        (5,): -0.0945443578134895
        (6,): -0.18585628977962745
        (7,): -0.2890251265097397
)

Visualize the result

iv.plot_force(feature_names=feature_names)

References

[1]

Scott M. Lundberg and Su-In Lee. A unified approach to interpreting model predictions. In Advances in Neural Information Processing Systems 30: Annual Conference on Neural Information Processing Systems 2017, December 4-9, 2017, Long Beach, CA, USA, 4765–4774. 2017. URL: https://proceedings.neurips.cc/paper/2017/hash/8a20a8621978632d76c43dfd28b67767-Abstract.html.