HumRRO’s innovative spirit and exceptional technical expertise allowed us to serve as the ideal organizer and host of a machine learning competition for the Society of Industrial and Organizational Psychology (SIOP). The goal of the competition was to provide a novel mechanism for educating SIOP members about advances in predictive modeling. Our very own, Dr. Dan Putka spearheaded the special session as the first of its kind at the 2018 SIOP Conference in Chicago. Additionally, the competition provided the data sponsor, Eli Lilly, with valuable, crowd-sourced insight into a prediction problem of interest to their organization.
Twelve different teams, including a HumRRO team firewalled from the organizers, participated in this year’s competition representing 14 different institutions across academe and industry. Teams participating in the competition were tasked with building predictive models of voluntary employee turnover based on de-identified data on more than 150 variables from over 30,000 employees. Over 180 entries (i.e., models) were submitted over the course of the four-month competition, which culminated with the top four teams describing their best performing models at a special session at the 2018 SIOP Conference. As a true testament to our modeling capabilities, HumRRO’s team was one of best, allowing us to present during the session.
To learn more about the competition, download our overview slides from the 2018 SIOP special session.