Conditional expectation network for SHAP

A very popular model-agnostic technique for explaining predictive models is the SHapley Additive exPlanation (SHAP). The two most popular versions of SHAP are a conditional expectation version and an unconditional expectation version (the latter is also known as interventional SHAP). Except for tree-based methods, usually the unconditional version is used (for computational reasons). We provide a (surrogate) neural network approach which allows us to efficiently calculate the conditional version for both neural networks and other regression models, and which properly considers the dependence structure in the feature components. This proposal is also useful to provide drop1 and anova analyses in complex regression models which are similar to their generalized linear model (GLM) counterparts, and we provide a partial dependence plot (PDP) counterpart that considers the right dependence structure in the feature components.

[1]  Łukasz Delong,et al.  The use of autoencoders for training neural networks with mixed categorical and numerical features , 2023, ASTIN Bulletin.

[2]  Patrice Gaillardetz,et al.  Risk allocation through shapley decompositions, with applications to variable annuities , 2023, ASTIN Bulletin.

[3]  F. Lindskog,et al.  Local bias adjustment, duration-weighted probabilities, and automatic construction of tariff cells , 2023, SSRN Electronic Journal.

[4]  Yann Pequignot,et al.  Understanding Interventional TreeSHAP : How and Why it Works , 2022, ArXiv.

[5]  Marvin N. Wright,et al.  Unifying local and global model explanations by functional decomposition of low dimensional structures , 2022, AISTATS.

[6]  Christian Jonen,et al.  Neural networks meet least squares Monte Carlo at internal model data , 2022, European Actuarial Journal.

[7]  Pradeep Ravikumar,et al.  Faith-Shap: The Faithful Shapley Interaction Index , 2022, J. Mach. Learn. Res..

[8]  Ross B. Girshick,et al.  Masked Autoencoders Are Scalable Vision Learners , 2021, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR).

[9]  T. Gneiting,et al.  Regression Diagnostics meets Forecast Evaluation: Conditional Calibration, Reliability Diagrams, and Coefficient of Determination , 2021, 2108.03210.

[10]  Mario V. Wuthrich,et al.  Statistical Foundations of Actuarial Learning and its Applications , 2021, SSRN Electronic Journal.

[11]  Trevor Hastie,et al.  Causal Interpretations of Black-Box Models , 2019, Journal of business & economic statistics : a publication of the American Statistical Association.

[12]  Anne-Sophie Krah,et al.  Least-Squares Monte Carlo for Proxy Modeling in Life Insurance: Neural Networks , 2020, Risks.

[13]  Ronald Richman,et al.  AI in actuarial science – a review of recent advances – part 2 , 2020, Annals of Actuarial Science.

[14]  Ronald Richman,et al.  AI in actuarial science – a review of recent advances – part 1 , 2020, Annals of Actuarial Science.

[15]  Patrick Cheridito,et al.  Assessing Asset-Liability Risk with Neural Networks , 2020, Risks.

[16]  Mathias Lindholm,et al.  DISCRIMINATION-FREE INSURANCE PRICING , 2020, ASTIN Bulletin.

[17]  Hugh Chen,et al.  From local explanations to global understanding with explainable AI for trees , 2020, Nature Machine Intelligence.

[18]  Dominik Janzing,et al.  Feature relevance quantification in explainable AI: A causality problem , 2019, AISTATS.

[19]  Mukund Sundararajan,et al.  The many Shapley values for model explanation , 2019, ICML.

[20]  Kjersti Aas,et al.  Explaining individual predictions when features are dependent: More accurate approximations to Shapley values , 2019, Artif. Intell..

[21]  Scott Lundberg,et al.  A Unified Approach to Interpreting Model Predictions , 2017, NIPS.

[22]  D. Apley,et al.  Visualizing the effects of predictor variables in black box supervised learning models , 2016, Journal of the Royal Statistical Society: Series B (Statistical Methodology).

[23]  Cheng Guo,et al.  Entity Embeddings of Categorical Variables , 2016, ArXiv.

[24]  Fabrizio Durante,et al.  Computational Actuarial Science with R , 2015 .

[25]  Pascal Vincent,et al.  Artificial Neural Networks Applied to Taxi Destination Prediction , 2015, DC@PKDD/ECML.

[26]  Eric R. Ziegel,et al.  Generalized Linear Models , 2002, Technometrics.

[27]  J. Friedman.  Greedy function approximation: A gradient boosting machine. , 2001 .

[28]  John N. Tsitsiklis,et al.  Regression methods for pricing complex American-style options , 2001, IEEE Trans. Neural Networks.

[29]  Francis A. Longstaff,et al.  Valuing American Options by Simulation: A Simple Least-Squares Approach , 2001 .

[30]  J. Carriére.  Valuation of the early-exercise price for options using simulations and nonparametric regression , 1996 .

[31]  L. Shapley.  A Value for n-person Games , 1988 .

[32]  Mario V. Wuthrich,et al.  SHAP for Actuaries: Explain any Model , 2023, SSRN Electronic Journal.

[33]  R Core Team,et al.  R: A language and environment for statistical computing. , 2014 .

[34]  L. Breiman.  Random Forests , 2001, Machine Learning.