Deep Learning for Revenue-Optimal Auctions with Budgets
Deep Learning for Revenue-Optimal Auctions with Budgets
复制标题
DOI：
--
复制
发表时间：
2018-07
期刊：
影响因子：
--
通讯作者：
Zhe Feng;H. Narasimhan;D. Parkes
中科院分区：
文献类型：
--
作者：
Zhe Feng;H. Narasimhan;D. Parkes
为私人预算机构设计收益最大化的拍卖是一项艰巨的任务。即使是单件物品的情况也没有被完全理解，对于最优的、主导策略激励相容的、两件物品的拍卖也没有分析结果。在这项工作中，我们将机制建模为神经网络，并使用机器学习进行最优拍卖的自动设计。我们扩展了懊悔网框架来处理私人预算约束和贝叶斯激励兼容性。我们发现新的拍卖非常接近激励兼容性和高收入的私人预算多单位拍卖，包括单位需求投标人的问题。为了基准测试的目的，我们还说明了\em RegretNet可以在更简单的设置下获得本质上最优的设计，其中解析解可用~\citeCHE2000，Malakhov2008,PAI2014。
The design of revenue-maximizing auctions for settings with private budgets is a hard task. Even the single-item case is not fully understood, and there are no analytical results for optimal, dominant-strategy incentive compatibile, two-item auctions. In this work, we model a mechanism as a neural network, and use machine learning for the automated design of optimal auctions. We extend the \em RegretNet framework~\citedeep-auction to handle private budget constraints and Bayesian incentive compatibility. We discover new auctions with very close approximations to incentive-compatibility and high revenue for multi-unit auctions with private budgets, including problems with unit-demand bidders. For benchmarking purposes, we also illustrate that \em RegretNet can obtain essentially optimal designs for simpler settings where analytical solutions are available~\citeCHE2000,Malakhov2008,PAI2014.