EURO-Online login
- New to EURO? Create an account
- I forgot my username and/or my password.
- Help with cookies
(important for IE8 users)
1859. Regression Markets: Incentivizing Data Sharing for Forecasting
Invited abstract in session WA-35: Data Valuation from Data-driven Optimization, stream Stochastic, Robust and Distributionally Robust Optimization.
Wednesday, 8:30-10:00Room: 44 (building: 303A)
Authors (first author is the speaker)
| 1. | Thomas Falconer
|
| Department of Wind and Energy Systems, Technical University of Denmark | |
| 2. | Jalal Kazempour
|
| Department of Wind and Energy Systems, Technical University of Denmark (DTU) | |
| 3. | Pierre Pinson
|
| Dyson School of Design Engineering, Imperial College London |
Abstract
Machine learning tasks rely heavily on the quality of input data, yet acquiring adequate datasets can often be challenging. Useful datasets are typically distributed amongst various owners who may, in practice, be competitors in downstream markets, making them reluctant to share information. In contrast to existing frameworks that address distributed and privacy-preserving (incentive-free) learning, we explore here a novel market-based framework, called regression markets, to provide financial incentives for data sharing. Agents aiming to improve their forecasts can post a regression task, for which others can contribute by sharing their data and get monetarily rewarded for it. We introduce the market design, provide its desirable properties, and discuss some of the open challenges with practical applications of treating data as a tradeable good.
Keywords
- Forecasting
- OR in Energy
- Game Theory
Status: accepted
Back to the list of papers