A Unified Model for Bid Landscape Forecasting in the Mixed Auction Types of Real-Time Bidding | IEEE Conference Publication | IEEE Xplore

A Unified Model for Bid Landscape Forecasting in the Mixed Auction Types of Real-Time Bidding


Abstract:

The increasing demand for online advertising leads to a strong competition in Real-Time Biddinng (RTB) industry. It requires Demand-Side Platforms (DSPs) to perform a pro...Show More

Abstract:

The increasing demand for online advertising leads to a strong competition in Real-Time Biddinng (RTB) industry. It requires Demand-Side Platforms (DSPs) to perform a proper market price modeling that predicts the landscape of competitors’ bids, in order to maximize their profits. Under this circumstance, RTB industry has recently been changing from second-price auctions (SPA) to first-price auctions (FPA), and thus DSPs now face two different auction types simultaneously. Most previous studies on market price modeling, however, have been suggested mainly for SPA, and the censorship problem of FPA has still been largely unexplored. Moreover, since those studies focused on only one auction type (either SPA or FPA), it takes additional computational and operational resources to apply these approaches to an environment where two types of auction are mixed. To this end, we introduce a novel unified approach named Conditional Distribution Modeling (CDM) to estimate market price probability distribution for SPA and FPA altogether. We utilize survival analysis and neural network to handle both right-censored problem in SPA and doubly-censored problem in FPA. Our model outperformed the previous models specifically developed either for SPA or FPA on two large-scale real-world datasets. Furthermore, our approach showed robust performance even when applied to mixed datasets with two auction types. These results indicate that our proposed model has an advantage in terms of both performance metrics and operational efficiency in a complex RTB environment.
Date of Conference: 17-20 December 2022
Date Added to IEEE Xplore: 26 January 2023
ISBN Information:
Conference Location: Osaka, Japan

References

References is not available for this document.