A new evidence sufficiency model with four dimensions and seven proxy categories enables monitoring of ML risk systems under delayed ground truth, detecting covariate and mixed drift but not concept drift without feature changes.
Title resolution pending
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 3verdicts
UNVERDICTED 3roles
background 1polarities
background 1representative citing papers
CM-DCM jointly models direct and delayed conversions in pre-promotion e-commerce via multi-task learning, personalized gating, and counterfactual transition probabilities from add-to-cart, outperforming baselines with gains in ad revenue and GMV in A/B tests.
Fairness-induced exploration in recommenders exhibits diminishing or non-monotonic returns that vary by user interaction history, with low-history users saturating sooner.
citing papers explorer
-
Evidence Sufficiency Under Delayed Ground Truth: Proxy Monitoring for Risk Decision Systems
A new evidence sufficiency model with four dimensions and seven proxy categories enables monitoring of ML risk systems under delayed ground truth, detecting covariate and mixed drift but not concept drift without feature changes.
-
Counterfactual Multi-task Learning for Delayed Conversion Modeling in E-commerce Sales Pre-Promotion
CM-DCM jointly models direct and delayed conversions in pre-promotion e-commerce via multi-task learning, personalized gating, and counterfactual transition probabilities from add-to-cart, outperforming baselines with gains in ad revenue and GMV in A/B tests.
-
Modeling User Exploration Saturation: When Recommender Systems Should Stop Pushing Novelty
Fairness-induced exploration in recommenders exhibits diminishing or non-monotonic returns that vary by user interaction history, with low-history users saturating sooner.