The University of Iowa’s Statistics and Actuarial Science graduate students, Behrooz Khalil Loo, Mahdi Saeidi Kousha, Nathan Munshower, David Roth and Nikita Jaiswal were one of the top teams at the 2024 Travelers University Modeling Competition that took place on Kaggle. They shared the following regarding the competition: “The competition was held on Kaggle and we heard about it in the course STAT:5400 Computing in Statistics taught by Dr. Wang”. The goal of the competition was to build a model to forecast the number of times a policyholder is likely to call based on historical that aims to optimize resource allocation and enhance cost-efficiency in call center operations. The call count is zero-inflated and highly right skewed. Relative Gini index was utilized to evaluate the model. Teams with a higher score than the benchmark can move to the presentation stage. They presented their prediction and analysis results as well as business insights. They expected to make real world connections to the model results. Their team tied for 1st out of 22 in Kaggle.   The final models used for this competition were the “Gradient Tree-Boosted Tweedie Hurdle Model”

Congratulations to "The Traveling Salesmen"​ team on their accomplishment!