Online Prediction via Continuous Artificial Prediction Markets

Fatemeh Jahedpari, Talal Rahwan, Sattar Hashemi, Tomasz P Michalak, Marina De Vos, Julian Padget, Wei Lee Woon

Research output: Contribution to journalArticle

Abstract

Prediction markets are well-established tools for aggregating information from diverse sources into accurate forecasts. Their success has been demonstrated in a wide range applications, including presidential campaigns, sporting events, and economic outcomes. Recently, they've been introduced to the machine learning community in the form of artificial prediction markets, in which algorithms trade contracts reflecting their levels of confidence. To date, these markets have mostly been studied in the context of offline classification, with promising results. The authors extend them to enable their use in online regression and introduce adaptive trading strategies informed by individual trading history and the ability of participants to revise their predictions by reflecting on the wisdom of the crowd, which is manifested in the collective performance of the market. The authors empirically evaluate their model using multiple datasets and show that it outperforms several well-established techniques from the literature on online regression.
LanguageEnglish
Article number7851146
Pages61-68
JournalIEEE Intelligent Systems
Volume32
Issue number1
DOIs
StatusPublished - 13 Feb 2017

Fingerprint

Learning systems
History
Economics

Cite this

Jahedpari, F., Rahwan, T., Hashemi, S., Michalak, T. P., De Vos, M., Padget, J., & Woon, W. L. (2017). Online Prediction via Continuous Artificial Prediction Markets. DOI: 10.1109/MIS.2017.12

Online Prediction via Continuous Artificial Prediction Markets. / Jahedpari, Fatemeh; Rahwan, Talal; Hashemi, Sattar; Michalak, Tomasz P; De Vos, Marina; Padget, Julian; Woon, Wei Lee.

In: IEEE Intelligent Systems, Vol. 32 , No. 1, 7851146, 13.02.2017, p. 61-68.

Research output: Contribution to journalArticle

Jahedpari, F, Rahwan, T, Hashemi, S, Michalak, TP, De Vos, M, Padget, J & Woon, WL 2017, 'Online Prediction via Continuous Artificial Prediction Markets' IEEE Intelligent Systems, vol. 32 , no. 1, 7851146, pp. 61-68. DOI: 10.1109/MIS.2017.12
Jahedpari F, Rahwan T, Hashemi S, Michalak TP, De Vos M, Padget J et al. Online Prediction via Continuous Artificial Prediction Markets. IEEE Intelligent Systems. 2017 Feb 13;32 (1):61-68. 7851146. Available from, DOI: 10.1109/MIS.2017.12
Jahedpari, Fatemeh ; Rahwan, Talal ; Hashemi, Sattar ; Michalak, Tomasz P ; De Vos, Marina ; Padget, Julian ; Woon, Wei Lee. / Online Prediction via Continuous Artificial Prediction Markets. In: IEEE Intelligent Systems. 2017 ; Vol. 32 , No. 1. pp. 61-68
@article{b8b6e8c3ed38417d928d232be95cdc3a,
title = "Online Prediction via Continuous Artificial Prediction Markets",
abstract = "Prediction markets are well-established tools for aggregating information from diverse sources into accurate forecasts. Their success has been demonstrated in a wide range applications, including presidential campaigns, sporting events, and economic outcomes. Recently, they've been introduced to the machine learning community in the form of artificial prediction markets, in which algorithms trade contracts reflecting their levels of confidence. To date, these markets have mostly been studied in the context of offline classification, with promising results. The authors extend them to enable their use in online regression and introduce adaptive trading strategies informed by individual trading history and the ability of participants to revise their predictions by reflecting on the wisdom of the crowd, which is manifested in the collective performance of the market. The authors empirically evaluate their model using multiple datasets and show that it outperforms several well-established techniques from the literature on online regression.",
author = "Fatemeh Jahedpari and Talal Rahwan and Sattar Hashemi and Michalak, {Tomasz P} and {De Vos}, Marina and Julian Padget and Woon, {Wei Lee}",
year = "2017",
month = "2",
day = "13",
doi = "10.1109/MIS.2017.12",
language = "English",
volume = "32",
pages = "61--68",
journal = "IEEE Intelligent Systems",
issn = "1541-1672",
publisher = "IEEE",
number = "1",

}

TY - JOUR

T1 - Online Prediction via Continuous Artificial Prediction Markets

AU - Jahedpari,Fatemeh

AU - Rahwan,Talal

AU - Hashemi,Sattar

AU - Michalak,Tomasz P

AU - De Vos,Marina

AU - Padget,Julian

AU - Woon,Wei Lee

PY - 2017/2/13

Y1 - 2017/2/13

N2 - Prediction markets are well-established tools for aggregating information from diverse sources into accurate forecasts. Their success has been demonstrated in a wide range applications, including presidential campaigns, sporting events, and economic outcomes. Recently, they've been introduced to the machine learning community in the form of artificial prediction markets, in which algorithms trade contracts reflecting their levels of confidence. To date, these markets have mostly been studied in the context of offline classification, with promising results. The authors extend them to enable their use in online regression and introduce adaptive trading strategies informed by individual trading history and the ability of participants to revise their predictions by reflecting on the wisdom of the crowd, which is manifested in the collective performance of the market. The authors empirically evaluate their model using multiple datasets and show that it outperforms several well-established techniques from the literature on online regression.

AB - Prediction markets are well-established tools for aggregating information from diverse sources into accurate forecasts. Their success has been demonstrated in a wide range applications, including presidential campaigns, sporting events, and economic outcomes. Recently, they've been introduced to the machine learning community in the form of artificial prediction markets, in which algorithms trade contracts reflecting their levels of confidence. To date, these markets have mostly been studied in the context of offline classification, with promising results. The authors extend them to enable their use in online regression and introduce adaptive trading strategies informed by individual trading history and the ability of participants to revise their predictions by reflecting on the wisdom of the crowd, which is manifested in the collective performance of the market. The authors empirically evaluate their model using multiple datasets and show that it outperforms several well-established techniques from the literature on online regression.

UR - http://dx.doi.org/10.1109/MIS.2017.12

U2 - 10.1109/MIS.2017.12

DO - 10.1109/MIS.2017.12

M3 - Article

VL - 32

SP - 61

EP - 68

JO - IEEE Intelligent Systems

T2 - IEEE Intelligent Systems

JF - IEEE Intelligent Systems

SN - 1541-1672

IS - 1

M1 - 7851146

ER -