Learning-Augmented Online Bidding in Stochastic Settings
By: Spyros Angelopoulos, Bertrand Simon
Potential Business Impact:
Helps online auctions make smarter, fairer bids.
Online bidding is a classic optimization problem, with several applications in online decision-making, the design of interruptible systems, and the analysis of approximation algorithms. In this work, we study online bidding under learning-augmented settings that incorporate stochasticity, in either the prediction oracle or the algorithm itself. In the first part, we study bidding under distributional predictions, and find Pareto-optimal algorithms that offer the best-possible tradeoff between the consistency and the robustness of the algorithm. In the second part, we study the power and limitations of randomized bidding algorithms, by presenting upper and lower bounds on the consistency/robustness tradeoffs. Previous works focused predominantly on oracles that do not leverage stochastic information on the quality of the prediction, and deterministic algorithms.
Similar Papers
No-Regret Online Autobidding Algorithms in First-price Auctions
CS and Game Theory
Makes online ads pay better, even with rules.
Strategy-robust Online Learning in Contextual Pricing
CS and Game Theory
Sells things online even when buyers lie.
HALO: Hindsight-Augmented Learning for Online Auto-Bidding
Machine Learning (CS)
Helps online ads reach the right people cheaper.