Modeling Theater Attendance with Zero-Inflated Negative Binomial Regression by Ryan Crants ’24

The ZINBox Office – Modeling Theater Attendance with Zero-Inflated Negative Binomial Regression (ZINB) by Ryan Crants ’24, Wednesday April 3, 1:10 – 1:50pm, North Science Building 015, Wachenheim, Statistics Colloquium

Abstract: Zero-Inflated Negative Binomial (ZINB) regression is a statistical technique used to model situations in which data contains an extreme amount of zeroes—so many zeroes, in fact, that many of them have a cause outside the scope of the study! ZINB regression allows us to account for this—to try and figure out which zeroes are due to our parameters of interest and which are due to something else. We will motivate the model’s usage, go through the model itself and all of its moving parts, and then examine a use case: modeling attendance to live theatrical performances in a general population.