Modeling Elections with Google Search Data Using Gaussian Process Regression by Andrew Nachamkin ‘24

Statistics Colloquium: Modeling Elections with Google Search Data Using Gaussian Process Regression by Andrew Nachamkin ‘24, Wednesday February 21, 1:10 – 1:50pn, North Science Building 015, Wachenheim

Abstract: In this colloquium, we will first look at libraries in R and Python that allow us to download and parse Google Trends data. With this data, we can then use a machine learning technique called Gaussian Process Regression to try and model recent presidential elections, with specific attention on partisan search terms that can help predict which way each state may swing.