# UKD: Debiasing Conversion Rate Estimation via Uncertainty-regularized Knowledge Distillation

arXiv: https://arxiv.org/abs/2201.08024 · HTML: https://mlsanity.com/paper/2201.08024/

## Key facts

- Cited by 14 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Jan 2022.

## Built on by (most cited first, 14 of 14)

- [Unbiased Knowledge Distillation for Recommendation](https://mlsanity.com/paper/2211.14729/) (2022)
- [Posterior Probability Matters: Doubly-Adaptive Calibration for Neural Predictions in Online Advertising](https://mlsanity.com/paper/2205.07295/) (2022)
- [Capturing Conversion Rate Fluctuation during Sales Promotions: A Novel Historical Data Reuse Approach](https://mlsanity.com/paper/2305.12837/) (2023)
- [Rethinking Missing Data: Aleatoric Uncertainty-Aware Recommendation](https://mlsanity.com/paper/2209.11679/) (2022)
- [FedAds: A Benchmark for Privacy-Preserving CVR Estimation with Vertical Federated Learning](https://mlsanity.com/paper/2305.08328/) (2023)
- [Entire-Space Variational Information Exploitation for Post-Click Conversion Rate Prediction](https://mlsanity.com/paper/2502.15687/) (2024)
- [Uncertainty Calibration for Counterfactual Propensity Estimation in Recommendation](https://mlsanity.com/paper/2303.12973/) (2023)
- [Contrastive Learning for Conversion Rate Prediction](https://mlsanity.com/paper/2307.05974/) (2023)
- [A Model-based Multi-Agent Personalized Short-Video Recommender System](https://mlsanity.com/paper/2405.01847/) (2024)
- [Privacy Preserving Conversion Modeling in Data Clean Room](https://mlsanity.com/paper/2505.14959/) (2025)
- [BPL: Bias-adaptive Preference Distillation Learning for Recommender System](https://mlsanity.com/paper/2510.16076/) (2025)
- [Uncertainty Modeling for Multi-Objective RTA Interception with Distillation Acceleration](https://mlsanity.com/paper/2511.05582/) (2025)
- [Conversion rate prediction in online advertising: modeling techniques, performance evaluation and future directions](https://mlsanity.com/paper/2512.01171/) (2025)
- [AdFL: In-Browser Federated Learning for Online Advertisement](https://mlsanity.com/paper/2602.06336/) (2026)

## Abstract (arXiv)

In online advertising, conventional post-click conversion rate (CVR) estimation models are trained using clicked samples. However, during online serving the models need to estimate for all impression ads, leading to the sample selection bias (SSB) issue. Intuitively, providing reliable supervision signals for unclicked ads is a feasible way to alleviate the SSB issue. This paper proposes an uncertainty-regularized knowledge distillation (UKD) framework to debias CVR estimation via distilling knowledge from unclicked ads. A teacher model learns click-adaptive representations and produces pseudo-conversion labels on unclicked ads as supervision signals. Then a student model is trained on both clicked and unclicked ads with knowledge distillation, performing uncertainty modeling to alleviate the inherent noise in pseudo-labels. Experiments on billion-scale datasets show that UKD outperforms previous debiasing methods. Online results verify that UKD achieves significant improvements.