thanks for presenting! related to #10, i'm curious whether you've looked at all at using models with these features to predict voting behavior for elected officials throughout time? I'd suspect that certain dimensions wax and wane in their predictive power over time – using supervised methods like these (or some that preserve interpretability more directly) might offer a way of answering that question.
thanks for presenting! related to #10, i'm curious whether you've looked at all at using models with these features to predict voting behavior for elected officials throughout time? I'd suspect that certain dimensions wax and wane in their predictive power over time – using supervised methods like these (or some that preserve interpretability more directly) might offer a way of answering that question.