B.E.E.G.U.S.
The Bicameral Election Estimation for General elections in the United States
Mission
To create an AI that when given election data is able to predict the most likely winner.
Step One
Data Collection
Collect and format data to train AI. This data includes :
Census
Polling
Contribution
Report Cards
Step Two
Statistical Development
Create a variety of reliable statistical methodologies that uses input data to accurately predict the likely outcome of a given election.
Step Three
Algorithm Weighting
Create a sub-model which generates accurate predictions from latent-space output of statistical methodologies.
Step Four
Development
Develop net-based website to serve as customer portal to prediction services.
Questions?
Contact qjimenez@stevens.edu, ehannah@stevens.edu, econe@stevens.edu, or adimoski@stevens.edu to get more information on the project.