Persistent Customers

Authors

• Megan Chaffey, University of Nebraska - LincolnFollow
• Lauren Kasparek, University of Nebraska-LincolnFollow
• Cole Johnson, University of Nebraska - LincolnFollow
• Joshua Madsen, University of Nebraska - LincolnFollow
• Ryan Orth, University of Nebraska - LincolnFollow

Date of this Version

5-2023

Document Type

Thesis

Citation

Chaffey, M., Kasparek, L., Johnson, C., Madsen, J., and Orth, R. 2023. Persistent Customers. Undergraduate Honors Thesis. University of Nebraska-Lincoln.

Comments

Copyright Megan Chaffey, Lauren Kasparek, Cole Johnson, Joshua Madsen, and Ryan Orth 2023

Abstract

The goal of this project was to produce a predictive model that scores a policyholder’s likelihood to churn within a given time window. Mutual of Omaha understood policyholder persistence through an actuarial lens which is limited to variables that the industry understands well and to probability theories and statistical techniques that have a long history. Machine learning driven model discovery, by contrast, avails itself of a wider set of predictive variables and leverages the unusual size of datasets in modern enterprises to validate the patterns it discovers.