Globally-Optimal CTA Model of Voting for US Senators December 9, 2015January 8, 2019 ~ paulyarnold Paul R. Yarnold Optimal Data Analysis, LLC A globally-optimal classification tree analysis (GO-CTA) model yields relatively strong predictive accuracy in modeling the behavior of N = 2,843 people voting for Senators in statewide elections. View journal article Like this:Like Loading... Related Published by paulyarnold View all posts by paulyarnold