Sustainable Online Reinforcement Learning for Auto-bidding

TL;DR

This paper introduces a sustainable online reinforcement learning framework for auto-bidding in advertising, directly interacting with real systems to improve safety and effectiveness, addressing the gap between virtual and real environments.

Contribution

It proposes a novel online RL framework with a safe exploration policy and variance-suppressed conservative Q-learning, overcoming the offline-online discrepancy in auto-bidding.

Findings

Outperforms state-of-the-art auto-bidding algorithms in simulations

Demonstrates effectiveness in real-world advertising systems

Provides theoretical safety guarantees for the exploration policy

Abstract

Recently, auto-bidding technique has become an essential tool to increase the revenue of advertisers. Facing the complex and ever-changing bidding environments in the real-world advertising system (RAS), state-of-the-art auto-bidding policies usually leverage reinforcement learning (RL) algorithms to generate real-time bids on behalf of the advertisers. Due to safety concerns, it was believed that the RL training process can only be carried out in an offline virtual advertising system (VAS) that is built based on the historical data generated in the RAS. In this paper, we argue that there exists significant gaps between the VAS and RAS, making the RL training process suffer from the problem of inconsistency between online and offline (IBOO). Firstly, we formally define the IBOO and systematically analyze its causes and influences. Then, to avoid the IBOO, we propose a sustainable online RL (SORL) framework that trains the auto-bidding policy by directly interacting with the RAS, instead of learning in the VAS. Specifically, based on our proof of the Lipschitz smooth property of the Q function, we design a safe and efficient online exploration (SER) policy for continuously collecting data from the RAS. Meanwhile, we derive the theoretical lower bound on the safety of the SER policy. We also develop a variance-suppressed conservative Q-learning (V-CQL) method to effectively and stably learn the auto-bidding policy with the collected data. Finally, extensive simulated and real-world experiments validate the superiority of our approach over the state-of-the-art auto-bidding algorithm.