Primary References

Partial Dependence Plots (PDPs)

• Friedman, J.H. (2001). Greedy Function Approximation: A Gradient Boosting Machine. Annals of Statistics, 29(5), 1189–1232.

  Introduces PDPs as a method to interpret the effect of features in ensemble methods. Foundational academic paper for PDPs.

• Greenwell, B.M. (2017). pdp: An R Package for Constructing Partial Dependence Plots. The R Journal, 9(1), 421-436.

  A detailed description of PDP implementation and interpretation.

Individual Conditional Expectation (ICE)

• Goldstein, A., Kapelner, A., Bleich, J., & Pitkin, E. (2015). Peeking inside the black box: Visualizing statistical learning with plots of individual conditional expectation. Journal of Computational and Graphical Statistics, 24(1), 44-65.

Canonical ICE paper, introduces ICE plots as an extension to PDPs that can handle individual-level variation.

Shapely Additive Explanations (SHAP)

• Lundberg, S.M., & Lee, S.-I. (2017). A Unified Approach to Interpreting Model Predictions. NeurIPS 30.

Proposes the SHAP framework and connects Shapley values from cooperative game theory to model interpretability

• Strumbelj, E., & Kononenko, I. (2014). Explaining prediction models and individual predictions with feature contributions. Knowledge and Information Systems, 41(3), 647-665.

Early work establishing the game-theoretic approach to explaining individual predictions, which inspired SHAP

Accumulated Local Effects

• Apley, D.W., & Zhu, J. (2020). Visualizing the Effects of Predictor Variables in Black Box Supervised Learning Models. Journal of the Royal Statistical Society: Series B, 82(4), 1059–1086.

Introduces ALE plots as an improvement over PDPs for cases with correlated features.

Local Interpretable Model-agnostic Explanations (LIME)

• Ribeiro, M.T., Singh, S., & Guestrin, C. (2016). “Why Should I Trust You?”: Explaining the Predictions of Any Classifier. Proceedings of the 22nd ACM SIGKDD.

Original LIME paper introducing local surrogate modeling as a tool for explaining black-box predictions.

General Interpretability Frameworks

• Molnar, C. (2023). Interpretable Machine Learning. Online Book.

Comprehensive and open-access book in the subject of model interpretability. Covers PDPs, SHAP, ICE, ALE, LIME, and more in theory and code

• Hooker, G. (2007). Generalized Functional ANOVA Diagnostics for High-Dimensional Functions of Dependent Variables. Journal of Computational and Graphical Statistics: 16(3), 709=732.