Research map / Learning theory and algorithms / Combinatorial and search optimization
Deep Learning—Powered Iterative Combinatorial Auctions
Jakob Weissteiner, Sven Seuken
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| replace quadratic-kernel SVRs with DNNs for value prediction | ICA elicitation · LSVM | efficiency (%): 96.03 -> 97.74 | Improved |
| replace quadratic-kernel SVRs with DNNs | ICA elicitation · GSVM | efficiency (%): 98.85 -> 98.63 | No clear effect |
| replace quadratic-kernel SVRs with DNNs | ICA elicitation · MRVM | efficiency (%): 94.58 -> 95.04 | Improved |
| replace quadratic-kernel SVRs with DNNs | ICA elicitation · GSVM, LSVM, MRVM | revenue (%): 77.80 -> 67.81 (GSVM) | Hurt |
| replace quadratic SVR with DNN | value prediction · LSVM national bidder |T|=200 | MAE test: 25.44 -> 20.58 | Improved |
| replace quadratic SVR with DNN | value prediction · GSVM national bidder |T|=200 | MAE test: 0.12 -> 2.22 | Hurt |
| replace quadratic SVR with DNN | value prediction · LSVM national bidder |T|=50 | MAE test: 32.61 -> 29.90 | Improved |
| use multi-layer DNNs instead of one-hidden-layer network chosen for best prediction (ablation) | DNN-ICA · GSVM train | efficiency (%): 98.27% -> 98.94% | Improved |
| use larger/multi-layer architecture (R:[32,32] N:[10,10,10]) vs R:[100] N:[10] (ablation) | DNN-ICA · LSVM train | efficiency (%): 96.78 -> 97.74 | Improved |
| decrease number of initial reports c0 from 70 to 10 (ablation) | DNN-ICA · MRVM train | efficiency (%): 93.54 -> 95.00 | Improved |
| increase initial reports c0 from 30 to 40 (ablation) | DNN-ICA · LSVM train | efficiency (%): 96.87 -> 97.40 | Improved |
| use DNN instead of SVR | DNN-ICA vs SVR-ICA · GSVM | number of queries: 41.9 -> 53 | Hurt |
Closest papers by meaning
- Prices, Bids, Values: One ML-Powered Combinatorial Auction to Rule Them All (ICML 2025 · similarity 0.90)
- Machine Learning-Powered Combinatorial Clock Auction (AAAI 2024 · similarity 0.90)
- BundleFlow: Deep Menus for Combinatorial Auctions by Diffusion-Based Optimization (NeurIPS 2025 · similarity 0.88)
- Auctionformer: A Unified Deep Learning Algorithm for Solving Equilibrium Strategies in Auction Games (ICML 2024 · similarity 0.87)
- Learning Optimal Auctions with Correlated Value Distributions (AAAI 2025 · similarity 0.86)
- Increasing Revenue in Efficient Combinatorial Auctions by Learning to Generate Artificial Competition (AAAI 2025 · similarity 0.85)
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.
