AI Paper Atlas

Research map / Probabilistic and scientific modeling / Molecules and proteins

[Re] FOCUS: Flexible Optimizable Counterfactual Explanations for Tree Ensembles

Kyosuke Morita

NeurIPS 2023 · Probabilistic and scientific modeling › Molecules and proteins · approach: Prediction-powered and conformal inference

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
Use FOCUS instead of Feature Tweaking, Euclidean distance on Wine Quality AdaBoosttree ensemble counterfactual search · Wine Quality ABd_mean (Euclidean): 0.267 -> 0.188Improved
Use FOCUS instead of Feature Tweaking, Euclidean distance on Shopping AdaBoosttree ensemble counterfactual search · Shopping ABd_mean (Euclidean): 0.126 -> 0.028Improved
Use FOCUS instead of Feature Tweaking on single Decision Trees, Euclidean distancetree counterfactual search · HELOC DTd_mean (Euclidean): 0.120 -> 0.133Hurt
Use FOCUS instead of Feature Tweaking on Decision Trees, Euclideantree counterfactual search · Shopping DTd_mean (Euclidean): 0.119 -> 0.142Hurt
Use FOCUS instead of Feature Tweaking, Euclidean on Wine Quality Decision Treetree counterfactual search · Wine Quality DTd_mean (Euclidean): 0.269 -> 0.268No clear effect
Use FOCUS instead of Feature Tweaking, Cosine distance on Wine Quality Decision Treetree counterfactual search · Wine Quality DTd_mean (Cosine): 0.030 -> 0.003Improved
Use FOCUS instead of Feature Tweaking with Manhattan distance on Random Forestsrandom forest counterfactual search · Wine Quality RFd_mean (Manhattan): 0.223 -> 0.312Hurt
Use FOCUS instead of Feature Tweaking with Manhattan distance on Shopping AdaBoostAdaBoost counterfactual search · Shopping ABd_mean (Manhattan): 0.142 -> 0.046Improved
Compare FOCUS and FT across 36 settings (4 datasets x 3 models x 3 distances)FOCUS · all datasetsd_mean: 20 settings FOCUS better, 8 FT better, 8 no differenceMixed
Switch from tree ensembles to single Decision Trees (model type)FOCUS vs FT · all datasetsd_meanMixed
Use FOCUS instead of DACE with Mahalanobis distance on Wine Quality DTcounterfactual search · Wine Quality DTd_mean (Mahalanobis): 1.325 -> 0.542Improved
Use FOCUS instead of DACE with Mahalanobis on Shopping AdaBoostcounterfactual search · Shopping ABd_mean (Mahalanobis): 3.230 -> 0.303Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy