The DMA Analytics Council organizes an annual analytics challenge, which is open to academia and industry, and draws participation from hundreds of teams from 50+ countries. This year's challenge will be sponsored by Cleveland Clinic, and will require the participants to solve a patient re-activation problem. Jigyasa Analytics will run the competition on behalf of the DMA Analytics Council. NORC, University of Chicago will judge the most innovative entry at the competition. We thank Epsilon for their support in providing zipcode level demographic and lifestyle data. We also thank University of Wisconsin, Public Health Institute for allowing us to use the County Health Ranking data.
This year’s analytics challenge involves identifying the patients most likely to respond to a re-activation campaign. Cleveland Clinic uses direct mail campaigns to win back patients who have in the past received treatment at the clinic for chronic conditions, but have not returned. These campaigns are administered periodically, and Cleveland Clinic would like to optimize the returns from the program.
Teams participating in the competition will be provided with solicitation and response data from past campaigns. They will also be provided with patient demographics, treatment history and information about referring physicians. In addition to the data provided by Cleveland Clinic, the participants will also have access to zipcode level demographic and lifestyle variables provided by Epsilon.
Participating teams will be required to build a predictive model that improves the efficiency of the patient re-activation program. The exact metric to be optimized will be announced when the data is released on August 05, 2013.
Teams will be provided with a scoring sample which can be used to test the performance of their models, by uploading the scored sample to the competition website. This sample is for testing purposes only and final evaluation will be based on a different sample which will only be shared towards the end of the competition.
In addition, NORC at the University of Chicago will evaluate the top 10 entries to identify the most innovative approach. Additional details about the innovation award will be made available at the time of the data release on August 05, 2013
Three finalists (top two based on the model performance, and the innovation award winner), will be invited to present their work at Annual DMA Conference at Chicago on October 14, 2013. The winners will be announced at the conference.