Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
157 tokens/sec
GPT-4o
43 tokens/sec
Gemini 2.5 Pro Pro
43 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

Shapley Curves: A Smoothing Perspective (2211.13289v5)

Published 23 Nov 2022 in stat.ML, cs.LG, and stat.ME

Abstract: This paper fills the limited statistical understanding of Shapley values as a variable importance measure from a nonparametric (or smoothing) perspective. We introduce population-level \textit{Shapley curves} to measure the true variable importance, determined by the conditional expectation function and the distribution of covariates. Having defined the estimand, we derive minimax convergence rates and asymptotic normality under general conditions for the two leading estimation strategies. For finite sample inference, we propose a novel version of the wild bootstrap procedure tailored for capturing lower-order terms in the estimation of Shapley curves. Numerical studies confirm our theoretical findings, and an empirical application analyzes the determining factors of vehicle prices.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (48)
  1. Explaining individual predictions when features are dependent: More accurate approximations to Shapley values. Artificial Intelligence, 298:103502.
  2. Explaining predictive models using Shapley values and non-parametric vine copulas. Dependence Modeling, 9(1):62–81.
  3. SHAFF: Fast and consistent SHApley eFfect estimates via random Forests. In International Conference on Artificial Intelligence and Statistics, pages 5563–5582. PMLR.
  4. Mean decrease accuracy for random forests: inconsistency, and a practical solution via the Sobol-MDA. Biometrika.
  5. Generalized Hoeffding-Sobol decomposition for dependent variables-application to sensitivity analysis. Electronic Journal of Statistics, 6:2420–2448.
  6. Algorithms to estimate Shapley value feature attributions. Nature Machine Intelligence, pages 1–12.
  7. Improving kernelshap: Practical Shapley value estimation using linear regression. In International Conference on Artificial Intelligence and Statistics, pages 3457–3465. PMLR.
  8. Explaining by removing: A unified framework for model explanation. Journal of Machine Learning Research, 22(209):1–90.
  9. Understanding global feature contributions with additive importance measures. Advances in Neural Information Processing Systems, 33:17212–17223.
  10. Dagenais, M. G. (1969). A threshold regression model. Econometrica, 37(2):193–203.
  11. Multiple regimes and cross-country growth behaviour. Journal of Applied Econometrics, 10(4):365–384.
  12. Fan, J. (1993). Local linear regression smoothers and their minimax efficiencies. The Annals of Statistics, pages 196–216.
  13. Direct estimation of low-dimensional components in additive models. The Annals of Statistics, 26(3):943–971.
  14. Shapley explainability on the data manifold. In International Conference on Learning Representations.
  15. Shapley value confidence intervals for attributing variance explained. Frontiers in Applied Mathematics and Statistics, 6:587199.
  16. Hansen, B. E. (2000). Sample splitting and threshold estimation. Econometrica, 68(3):575–603.
  17. Comparing nonparametric versus parametric regression fits. The Annals of Statistics, pages 1926–1947.
  18. Bootstrap simultaneous error bars for nonparametric regression. The Annals of Statistics, pages 778–796.
  19. Random planted forest: a directly interpretable tree ensemble. arXiv preprint arXiv:2012.14563.
  20. Unifying local and global model explanations by functional decomposition of low dimensional structures. In International Conference on Artificial Intelligence and Statistics, pages 7040–7060. PMLR.
  21. Fastshap: Real-time shapley value estimation. ICLR 2022.
  22. Inferring feature importance with uncertainties in high-dimensional data. arXiv preprint arXiv:2109.00855.
  23. On the robustness of removal-based feature attributions. Advances in Neural Information Processing Systems, 36.
  24. A kernel method of estimating structured nonparametric regression based on marginal integration. Biometrika, pages 93–100.
  25. From local explanations to global understanding with explainable AI for trees. Nature Machine Intelligence, 2(1):56–67.
  26. A unified approach to interpreting model predictions. Advances in Neural Information Processing Systems, 30.
  27. Mammen, E. (1992). Bootstrap, wild bootstrap, and asymptotic normality. Probability Theory and Related Fields, 93(4):439–455.
  28. Mammen, E. (1993). Bootstrap and wild bootstrap for high dimensional linear models. The Annals of Statistics, 21(1):255–285.
  29. Explainable AI for a no-teardown vehicle component cost estimation: A top-down approach. IEEE Transactions on Artificial Intelligence, 2(2):185–199.
  30. Beyond treeSHAP: Efficient computation of any-order Shapley interactions for tree ensembles. arXiv preprint arXiv:2401.12069.
  31. An optimization interpretation of integration and back-fitting estimators for separable nonparametric models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 60(1):217–222.
  32. A multilinear sampling algorithm to estimate Shapley values. In 2020 25th International Conference on Pattern Recognition (ICPR), pages 7992–7999. IEEE.
  33. Owen, A. B. (2014). Sobol’indices and Shapley value. SIAM/ASA Journal on Uncertainty Quantification, 2(1):245–251.
  34. On Shapley value for measuring importance of dependent inputs. SIAM/ASA Journal on Uncertainty Quantification, 5(1):986–1002.
  35. Petrov, V. V. (1975). Chapter v. In Sums of Independent Random Variables. Springer, New York.
  36. Multivariate locally weighted least squares regression. The Annals of Statistics, pages 1346–1370.
  37. Schmidt-Hieber, J. (2020). Nonparametric regression using deep neural networks with ReLU activation function. The Annals of Statistics, 48(4):1875–1897.
  38. Scornet, E. (2023). Trees, forests, and impurity-based variable importance in regression. Annales de l’Institut Henri Poincare (B) Probabilites et statistiques, 59(1):21–52.
  39. Consistency of random forests. The Annals of Statistics, 43(4):1716 – 1741.
  40. Shapley, L. S. (1953). A value for n-person games. Contributions to the Theory of Games, 28(2):307–317.
  41. Shapley effects for global sensitivity analysis: Theory and computation. SIAM/ASA Journal on Uncertainty Quantification, 4(1):1060–1083.
  42. Nonparametric estimation and testing of interaction in additive models. Econometric Theory, 18(2):197–251.
  43. Stone, C. J. (1982). Optimal global rates of convergence for nonparametric regression. The Annals of Statistics, pages 1040–1053.
  44. Stone, C. J. (1985). Additive regression and other nonparametric models. The Annals of Statistics, pages 689–705.
  45. Explaining instance classifications with interactions of subsets of feature values. Data & Knowledge Engineering, 68(10):886–904.
  46. The many Shapley values for model explanation. In International Conference on Machine Learning, pages 9269–9278. PMLR.
  47. Nonparametric identification of nonlinear time series: projections. Journal of the American Statistical Association, 89(428):1398–1409.
  48. Efficient nonparametric statistical inference on population feature importance using Shapley values. In International Conference on Machine Learning, pages 10282–10291. PMLR.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now
X Twitter Logo Streamline Icon: https://streamlinehq.com

Tweets