Research map / Learning theory and algorithms / Submodular and game algorithms
Selling Data To a Machine Learner: Pricing via Costly Signaling
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 |
|---|---|---|---|
| Sell entire dataset directly without sharing any data (no costly signaling) under multiplicatively separable valuation u(r,b)=f(b)h(r) | homogeneous data pricing · theoretical analysis (homogeneous data) | seller revenue | No clear effect |
| Share a subset of data for costly signaling before pricing (versus selling entire dataset directly) | homogeneous data pricing with non-separable valuation · Example 3.3 (homogeneous data) | seller revenue | Mixed |
| Use robust mechanism with prior mismatch epsilon between seller and learner priors | homogeneous data pricing · theoretical analysis (robust mechanism) | revenue lower bound relative to OPT: at least OPT - 4ϵ/(1-ϵ²)·ū | No clear effect |
| Cardinality constraint on shared features with costly signaling for heterogeneous data | heterogeneous data pricing · theoretical (NP-hardness) | approximation ratio of optimal mechanism | Hurt |
| Sell entire dataset directly without sharing features (simple algorithm) for heterogeneous data | heterogeneous data pricing · theoretical (approximation guarantee) | fraction of optimal revenue guaranteed: at least 1/k OPT, k = number of private types | No clear effect |
| Greedy feature selection maximizing revenue increase for heterogeneous data | heterogeneous data pricing · theoretical (Lemma H.2) | revenue increase approximation | Hurt |
Closest papers by meaning
- Non-Linear Pricing Restores Tractability for a Data Seller (NeurIPS 2026 · similarity 0.88)
- An Instrumental Value for Data Production and its Application to Data Pricing (ICML 2025 · similarity 0.86)
- Learning to Price Homogeneous Data (NeurIPS 2024 · similarity 0.86)
- Fairshare Data Pricing via Data Valuation for Large Language Models (NeurIPS 2025 · similarity 0.86)
- Safely Learning Optimal Auctions: A Testable Learning Framework for Mechanism Design (ICML 2025 · similarity 0.85)
- Learning Classifiers That Induce Markets (ICML 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.