| Name: |
Good Judgement Open Consensus Forecast |
| Abbreviation: |
GJOCon
|
| Owner: |
allicodi
|
| Team name: |
Computational Uncertainty Laboratory |
| Description: |
Good Judgement Open is a non-profit organization as a result of a four-year, $20 million US Government research project aimed at
improving the accuracy and timeliness of forecasts. Experts- defined as people with several years of experience in the study or modeling of infectious disease and/or public health decision making- are posed questions related to the ongoing COVID-19 outbreak using this platform. Predictive densities over potential future values from each member of the crowd are collected and aggregated into a single consensus forecast. |
| Contributors: |
Thomas McAndrew (Lehigh University) <mcandrew@lehigh.edu>, David Braun (Lehigh University), Eva Chen (Good Judgement Open) <chen@goodjudgment.com>, Luis Enrique Urtubey (Good Judgement Open) <decesaris@goodjudgment.com>, Damon Luk (Lehigh University), Allison Codi (Lehigh University) |
| License: |
Creative Commons Attribution 4.0 |
| Notes: |
|
| Citation: |
|
| Methods: |
Each month, we (i) pose questions to a crowd of subject matter experts and trained generalist forecasters, (ii) collect predictive densities over potential future values from each member of the crowd, and (iii) we aggregate this set of predictive densities into a single consensus forecast. We expect predictions from subject matter experts and trained forecasters to be accurate and calibrated because they have access to structured data, the same data computational models use, and because they have access to subjective, unstructured data often unavailable to computational models. |
| Home: |
https://github.com/computationalUncertaintyLab/aggStatModelsAndHumanJudgment_PUBL
|
| Auxiliary data: |
(No URL)
|