Estimating Quality of Approximated Shapley Values Using Conformal Prediction

Amr Alkhatib, Henrik Boström, Ulf Johansson
Proceedings of the Thirteenth Symposium on Conformal and Probabilistic Prediction with Applications, PMLR 230:158-174, 2024.

Abstract

Thanks to their theoretically proven properties, Shapley values have received a lot of attention as a means to explain predictions within the area of explainable machine learning. However, the computation of Shapley values is time-consuming and computationally expensive, in particular for datasets with high dimensionality, often rendering them impractical for generating timely explanations. Methods to approximate Shapley values, e.g., FastSHAP, offer a solution with adequate computational cost. However, such approximations come with a degree of uncertainty. Therefore, we propose a method to measure the fidelity of Shapley value approximations and use the conformal prediction framework to provide validity guarantees for the whole explanation in contrast to an earlier approach that offered validity guarantees on a per-feature importance basis, disregarding the relative importance of the remaining feature scores within the same explanation. We propose a set of difficulty estimation functions devised to consider the difficulty of explanation approximations. We provide a large-scale empirical investigation where the proposed difficulty estimators are evaluated with respect to their efficiency (interval size) in measuring the similarity to the ground truth Shapley values. The results suggest that the proposed approach can provide predictions coupled with informative validity guarantees (tight intervals), allowing the user to trust/reject the provided explanations based on their similarity to the ground truth values.

Cite this Paper

BibTeX
@InProceedings{pmlr-v230-alkhatib24a, title = {Estimating Quality of Approximated Shapley Values Using Conformal Prediction}, author = {Alkhatib, Amr and Bostr\"{o}m, Henrik and Johansson, Ulf}, booktitle = {Proceedings of the Thirteenth Symposium on Conformal and Probabilistic Prediction with Applications}, pages = {158--174}, year = {2024}, editor = {Vantini, Simone and Fontana, Matteo and Solari, Aldo and Boström, Henrik and Carlsson, Lars}, volume = {230}, series = {Proceedings of Machine Learning Research}, month = {09--11 Sep}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v230/main/assets/alkhatib24a/alkhatib24a.pdf}, url = {https://proceedings.mlr.press/v230/alkhatib24a.html}, abstract = {Thanks to their theoretically proven properties, Shapley values have received a lot of attention as a means to explain predictions within the area of explainable machine learning. However, the computation of Shapley values is time-consuming and computationally expensive, in particular for datasets with high dimensionality, often rendering them impractical for generating timely explanations. Methods to approximate Shapley values, e.g., FastSHAP, offer a solution with adequate computational cost. However, such approximations come with a degree of uncertainty. Therefore, we propose a method to measure the fidelity of Shapley value approximations and use the conformal prediction framework to provide validity guarantees for the whole explanation in contrast to an earlier approach that offered validity guarantees on a per-feature importance basis, disregarding the relative importance of the remaining feature scores within the same explanation. We propose a set of difficulty estimation functions devised to consider the difficulty of explanation approximations. We provide a large-scale empirical investigation where the proposed difficulty estimators are evaluated with respect to their efficiency (interval size) in measuring the similarity to the ground truth Shapley values. The results suggest that the proposed approach can provide predictions coupled with informative validity guarantees (tight intervals), allowing the user to trust/reject the provided explanations based on their similarity to the ground truth values.} }
Endnote
%0 Conference Paper %T Estimating Quality of Approximated Shapley Values Using Conformal Prediction %A Amr Alkhatib %A Henrik Boström %A Ulf Johansson %B Proceedings of the Thirteenth Symposium on Conformal and Probabilistic Prediction with Applications %C Proceedings of Machine Learning Research %D 2024 %E Simone Vantini %E Matteo Fontana %E Aldo Solari %E Henrik Boström %E Lars Carlsson %F pmlr-v230-alkhatib24a %I PMLR %P 158--174 %U https://proceedings.mlr.press/v230/alkhatib24a.html %V 230 %X Thanks to their theoretically proven properties, Shapley values have received a lot of attention as a means to explain predictions within the area of explainable machine learning. However, the computation of Shapley values is time-consuming and computationally expensive, in particular for datasets with high dimensionality, often rendering them impractical for generating timely explanations. Methods to approximate Shapley values, e.g., FastSHAP, offer a solution with adequate computational cost. However, such approximations come with a degree of uncertainty. Therefore, we propose a method to measure the fidelity of Shapley value approximations and use the conformal prediction framework to provide validity guarantees for the whole explanation in contrast to an earlier approach that offered validity guarantees on a per-feature importance basis, disregarding the relative importance of the remaining feature scores within the same explanation. We propose a set of difficulty estimation functions devised to consider the difficulty of explanation approximations. We provide a large-scale empirical investigation where the proposed difficulty estimators are evaluated with respect to their efficiency (interval size) in measuring the similarity to the ground truth Shapley values. The results suggest that the proposed approach can provide predictions coupled with informative validity guarantees (tight intervals), allowing the user to trust/reject the provided explanations based on their similarity to the ground truth values.
APA
Alkhatib, A., Boström, H. & Johansson, U.. (2024). Estimating Quality of Approximated Shapley Values Using Conformal Prediction. Proceedings of the Thirteenth Symposium on Conformal and Probabilistic Prediction with Applications, in Proceedings of Machine Learning Research 230:158-174 Available from https://proceedings.mlr.press/v230/alkhatib24a.html.

Related Material