No-Regret Online Autobidding Algorithms in First-price Auctions
By: Yuan Deng , Yilin Li , Wei Tang and more
Potential Business Impact:
Makes online ads pay better, even with rules.
Automated bidding to optimize online advertising with various constraints, e.g. ROI constraints and budget constraints, is widely adopted by advertisers. A key challenge lies in designing algorithms for non-truthful mechanisms with ROI constraints. While prior work has addressed truthful auctions or non-truthful auctions with weaker benchmarks, this paper provides a significant improvement: We develop online bidding algorithms for repeated first-price auctions with ROI constraints, benchmarking against the optimal randomized strategy in hindsight. In the full feedback setting, where the maximum competing bid is observed, our algorithm achieves a near-optimal $\widetilde{O}(\sqrt{T})$ regret bound, and in the bandit feedback setting (where the bidder only observes whether the bidder wins each auction), our algorithm attains $\widetilde{O}(T^{3/4})$ regret bound.
Similar Papers
Online Bidding under RoS Constraints without Knowing the Value
Machine Learning (CS)
Helps online ads make more money smarter.
Learning to Bid in Non-Stationary Repeated First-Price Auctions
Machine Learning (CS)
Helps ad buyers win more online ads.
Learning-Augmented Online Bidding in Stochastic Settings
CS and Game Theory
Helps online auctions make smarter, fairer bids.