double robust learning

DataFunSummit

May 1, 2024 · Artificial Intelligence

Causal Solutions for Recommendation System Bias and Practical Applications

This article presents causal inference–based methods to address bias in recommendation systems, covering the transformation of recommendation problems into causal problems, selection bias mitigation through double‑robust and multi‑robust learning, individual treatment effect estimation, and a case study on attention bias in music recommendation.