AIGB: Generative Auto-bidding via Conditional Diffusion Modeling

TL;DR

This paper introduces DiffBid, a generative diffusion model for auto-bidding in online advertising, overcoming limitations of traditional RL methods by modeling entire bid trajectories to improve stability and performance.

Contribution

The paper presents DiffBid, a novel conditional diffusion approach for auto-bidding that captures trajectory correlations and enhances long-term decision stability.

Findings

Achieves 2.81% increase in GMV in real-world tests.

Attains 3.36% increase in ROI in online A/B testing.

Demonstrates superior stability over Markovian RL methods.

Abstract

Auto-bidding plays a crucial role in facilitating online advertising by automatically providing bids for advertisers. Reinforcement learning (RL) has gained popularity for auto-bidding. However, most current RL auto-bidding methods are modeled through the Markovian Decision Process (MDP), which assumes the Markovian state transition. This assumption restricts the ability to perform in long horizon scenarios and makes the model unstable when dealing with highly random online advertising environments. To tackle this issue, this paper introduces AI-Generated Bidding (AIGB), a novel paradigm for auto-bidding through generative modeling. In this paradigm, we propose DiffBid, a conditional diffusion modeling approach for bid generation. DiffBid directly models the correlation between the return and the entire trajectory, effectively avoiding error propagation across time steps in long horizons. Additionally, DiffBid offers a versatile approach for generating trajectories that maximize given targets while adhering to specific constraints. Extensive experiments conducted on the real-world dataset and online A/B test on Alibaba advertising platform demonstrate the effectiveness of DiffBid, achieving 2.81% increase in GMV and 3.36% increase in ROI.