近年来，机制设计在工业界已取得巨大成功。受此影响，机制设计也成为经济学与计算机科学交叉领域的核心研究课题之一。然而，尽管花费数十年的努力，学界在理论与实践方面仍然面临诸多挑战。本书提出了人工智能驱动的机制设计框架，以提供一种替代方法来处理目前机制设计理论与实践中的一些问题。该框架包含两个互相交互的抽象模型：智能体模型和机制模型。结合人工智能与机制设计，我们可以解决利用单一领域技术无法解决的问题。例如极大缩小机制搜索空间、构建更现实的买家模型、更好地平衡各类目标等。本书从多物品拍卖，动态拍卖，以及多目标拍卖三个场景入手，分析并说明该框架对理论与实践均有帮助。
本书可供从事人工智能驱动设计及相关领域的高校师生、研究人员及相关技术人员阅读参考。

Contents
Chapter 1 Introduction
1.1 Mechanism Design
1.1.1 Social Choice Function
1.1.2 Mechanism
1.1.3 Implementation
1.1.4 Revelation Principle
1.1.5 Efficient Mechanisms
1.2 Auctions
1.3 Why AI-Driven
1.3.1 Challenges in Auction Design
1.3.2 The AI-Driven Framework
1.4 Organization of the Book
References
Chapter 2 Multi-Dimensional Mechanism Design via AI-Driven Approaches
2.1 Recovering Optimal Mechanisms with Simple Neural Networks
2.1.1 Background
2.1.2 Setting
2.1.3 Revisiting the Naive Mechanism
2.1.4 Network Structure of MenuNet
2.1.5 Recovering Known Results
2.2 Discovering Unknown Optimal Mechanisms
2.2.1 Experiment Results
2.2.2 Theoretic Analysis and Formal Proofs
2.3 Performance
References
Chapter 3 Dynamic Mechanism Design via AI-Driven Approaches
3.1 Dynamic Cost-Per-Action Auctions with Ex-Post IR Guarantees
3.1.1 Background
3.1.2 Our Contributions
3.1.3 Related Works
3.1.4 Setting and Preliminaries
3.1.5 Mechanisms
3.1.6 Truthfulness and Implementation
3.1.7 Impossibility Result
3.2 Dynamic Reserve Pricing via Reinforcement Mechanism Design
3.2.1 Background
3.2.2 Settings and Preliminaries
3.2.3 Bidder Behavior Model
3.2.4 Dynamic Mechanism Design as Markov Decision Process
References
Chapter 4 Multi-Objective Mechanism Design via AI-Driven Approaches
4.1 Balancing Objectives through Approximation Analysis
4.1.1 Background
4.1.2 Settings and Preliminaries
4.1.3 Generalized Virtual-Efficient Mechanisms
4.1.4 Experiments
4.2 Balancing Objectives through Machine Learning
4.2.1 Background
4.2.2 Market Clearing Loss
4.2.3 Theoretical Guarantees
4.2.4 Empirical Evaluation
References
Chapter 5 Summary and Future Directions
References