I'm evaluating different model-agnostic methods on gradient boosting or random forest model.

For Shapley, specifically TreeSHAP, the complexity is O(TLD^2) according to Lundberg et al. 2018. T: #trees, L:max #leaves, D:depth

What would be the complexity of ICE (Individual conditional expectation) on the ensemble tree models? Is there any reference or paper mentioned about the algorithm complexity?




Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.