Election modeling

Thanks to the big data boom, modeling solutions to a problem using AI algorithms is now possible. One of the key areas where predictive analytics firms have been working is in the field of modeling election results.

Historically, election results were modeled using two key methods. Voters were asked before elections about whom they would vote for, and this method was termed an opinion poll. The other means was when voters were asked who they voted for when they exited the polling booths. This method was called exit polling. Exit polls were typically more accurate in predicting the election results than opinion polls.

The accuracy of exit polls has been challenged in recent times. In the past three Indian elections and the most recent Australian elections of 2018, exit polls have been inaccurate.

As predictive analytics and statistical modeling have become more mainstream, several economic indicators and their correlation to election results have...