Bidding

About: Bidding is a research topic. Over the lifetime, 15371 publications have been published within this topic receiving 294233 citations. The topic is also known as: competitive bidding.

Deep Reinforcement Learning for Sponsored Search Real-time Bidding

Abstract: Bidding optimization is one of the most critical problems in online advertising. Sponsored search (SS) auction, due to the randomness of user query behavior and platform nature, usually adopts keyword-level bidding strategies. In contrast, the display advertising (DA), as a relatively simpler scenario for auction, has taken advantage of real-time bidding (RTB) to boost the performance for advertisers. In this paper, we consider the RTB problem in sponsored search auction, named SS-RTB. SS-RTB has a much more complex dynamic environment, due to stochastic user query behavior and more complex bidding policies based on multiple keywords of an ad. Most previous methods for DA cannot be applied. We propose a reinforcement learning (RL) solution for handling the complex dynamic environment. Although some RL methods have been proposed for online advertising, they all fail to address the "environment changing'' problem: the state transition probabilities vary between two days. Motivated by the observation that auction sequences of two days share similar transition patterns at a proper aggregation level, we formulate a robust MDP model at hour-aggregation level of the auction data and propose a control-by-model framework for SS-RTB. Rather than generating bid prices directly, we decide a bidding model for impressions of each hour and perform real-time bidding accordingly. We also extend the method to handle the multi-agent problem. We deployed the SS-RTB system in the e-commerce search auction platform of Alibaba. Empirical experiments of offline evaluation and online A/B test demonstrate the effectiveness of our method.

Bribery and Favoritism by Auctioneers in Sealed Bid Auctions

Abstract: We consider a model of bribery and favoritism in a sealed-bid first-price procurement auction. The auctioneer can award the contract to a dishonest supplier at the low bid of an honest supplier. We examine the equilibrium bidding functions of both suppliers when it is common knowledge that the dishonest supplier can bribe the auctioneer. Both efficient and inefficient bribes can arise and the resulting allocative distortion differs from the distortions in a first-price auction or an optimal auction. The expected price paid by the buyer is generally higher with bribery, but when efficient bribes occur, there can be cases in which bribery results in a lower expected price. We also examine the incentives for cost-reducing investment by the suppliers and find that bribery results in a lower industry capacity than the social optimum. Finally, we examine upfront bribes in which the suppliers compete to be favored by the auctioneer and find that the stronger supplier will pay a larger bribe than the weaker supplier.

Ambiguity Aversion in First-Price Sealed-Bid Auctions

Abstract: Experiments on first-price sealed-bid auctions with independent private values have shown that submitted bids typically exceed Nash-equilibrium predictions for risk-neutral bidders. Existing bidding models explain this phenomenon by assuming that the bidders are risk-averse and capable of drawing complete and correct inferences about their winning probabilities. In this article, we use the Choquet expected utility (CEU) theory to demonstrate that the observed bidding behavior can also be attributed to ambiguity aversion which causes the bidders to underestimate their chances of winning the auction. Empirical support for CEU bidding models is given through an analysis of recent bidding data.