About this document
Top-Off-Policy Correction For A REINFORCE Recommender System by JWHC CORUMANA PROJECT is a document available to read on EtoBox.
This paper presents a method for improving recommendation systems using a top-K off-policy correction approach within a REINFORCE framework. It addresses the challenges of sparse data and biases in large-scale recommender systems by leveraging logged implicit feedback and proposing a novel correction mechanism for multiple item recommendations. The effectiveness of the proposed methods is demonstrated through simulations and live experiments on YouTube, highlighting their potential to enhance user satisfact
- Author
- JWHC CORUMANA PROJECT
- Language
- EN