AntTech
Jun 10, 2019 · Artificial Intelligence
Generative Adversarial User Model for Reinforcement Learning‑Based Recommendation Systems
This article presents a model‑based reinforcement learning framework for recommendation systems that uses a generative adversarial user model to simultaneously learn user behavior dynamics and reward functions, enabling efficient Cascading‑DQN policy learning and achieving superior long‑term user rewards and click‑through rates in experiments.
Cascading DQNGenerative Adversarial Networksartificial intelligence
0 likes · 9 min read