# Accelerated Convergence for Counterfactual Learning to Rank

arXiv: https://arxiv.org/abs/2005.10615 · HTML: https://mlsanity.com/paper/2005.10615/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 21 May 2020.

## Built on by (most cited first, 3 of 3)

- [Ranking over Regression for Bayesian Optimization and Molecule Selection](https://mlsanity.com/paper/2410.09290/) (2024)
- [Federated Unbiased Learning to Rank](https://mlsanity.com/paper/2105.04761/) (2021)
- [Learning to Re-rank with Constrained Meta-Optimal Transport](https://mlsanity.com/paper/2305.00319/) (2023)

## Abstract (arXiv)

Counterfactual Learning to Rank (LTR) algorithms learn a ranking model from logged user interactions, often collected using a production system. Employing such an offline learning approach has many benefits compared to an online one, but it is challenging as user feedback often contains high levels of bias. Unbiased LTR uses Inverse Propensity Scoring (IPS) to enable unbiased learning from logged user interactions. One of the major difficulties in applying Stochastic Gradient Descent (SGD) approaches to counterfactual learning problems is the large variance introduced by the propensity weights. In this paper we show that the convergence rate of SGD approaches with IPS-weighted gradients suffers from the large variance introduced by the IPS weights: convergence is slow, especially when there are large IPS weights. To overcome this limitation, we propose a novel learning algorithm, called CounterSample, that has provably better convergence than standard IPS-weighted gradient descent methods. We prove that CounterSample converges faster and complement our theoretical findings with empirical results by performing extensive experimentation in a number of biased LTR scenarios -- across optimizers, batch sizes, and different degrees of position bias.