Research map / Bandits and online learning / Online convex optimization and games
Leveraging the Hints: Adaptive Bidding in Repeated First-Price Auctions
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Provide the hint's accuracy interval in addition to its point prediction | Repeated first-price auction bidding with fixed private values · Fixed private value, L in [1, T] | minimax regret: C \sqrt{ \log T \cdot T^{\frac{1}{q+1}}\cdot L^{\frac{q}{q+1}}} vs. C \left(\log T\right)^{\frac{1}{2}} \left( T\cdot L | Helped |
| Use hint intervals rather than no hints | Binned exponential weighting on real-world auction data · Real-world auction data, private-value bins | cumulative reward improvement: 3.55% | Helped |
| Use single point hints rather than no hints | Binned exponential weighting on real-world auction data · Real-world auction data, private-value bins | cumulative reward improvement: 2.96% | Helped |
| Use hint intervals instead of single hints | Repeated first-price auction bidding · Varying private values | minimax regret: \widetilde{\Theta}\left(\min \left\{T^{\frac{1}{q+1}} L^{\frac{q}{q+1}}, \sqrt{T}\right\}\right) vs. \widetilde{\Theta}\ | Helped |
| Exploit that competing bids have only K support values | Repeated first-price auctions with hint intervals · K-supported competing bids | regret: O\left(\min \left\{\sqrt{\log T \cdot T^{\frac{1}{q+1}}\cdot L^{\frac{q}{q+1}} \cdot K}, T^{\frac{1}{q+1}}\cdot L^{\frac | Helped |
| Use single hints when private values vary | Repeated first-price auctions · Varying private values | minimax regret: \widetilde{\Theta}\left(\sqrt{T}\right) | No clear effect |
| Make the meta algorithm adaptive to unknown L and K | Meta-Expert Learning Algorithm · K-supported competing bids with hint intervals | regret upper bound: O\left(\sqrt{\log T\cdot 2K \cdot T^{\frac{1}{q+1}}\cdot L^{\frac{q}{q+1}}}\right) | Helped |
Closest papers by meaning
- No-Regret Online Autobidding Algorithms in First-price Auctions (NeurIPS 2025 · similarity 0.90)
- Simultaneous Optimization of Bid Shading and Internal Auction for Demand-Side Platforms (AAAI 2024 · similarity 0.88)
- Learning-Augmented Online Bidding in Stochastic Settings (NeurIPS 2025 · similarity 0.88)
- Maximizing the Success Probability of Policy Allocations in Online Systems (AAAI 2024 · similarity 0.88)
- Learning Safe Strategies for Value Maximizing Buyers in Uniform Price Auctions (ICML 2025 · similarity 0.87)
- Comparing Uniform Price and Discriminatory Multi-Unit Auctions through Regret Minimization (NeurIPS 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.