The Heritage Health Prize is now open! This data prediction competion, which offers $3.2M in prizes over the next two years to the team that can best predict health outcomes from anoymized data, is described in Forbes today:
Using predictive data modeling, contestants will examine three years of historical medical data from anonymized real-life people. The challenge is to create an algorithm that will predict how many days each person will spend in the hospital in the one year after that three-years of data. The winner will be the contestant that is closest on average to the actual number of hospital days for each patient.
The goal is to create an “early warning system” for managed care providers and provide a better way to identify which patients need care immediately to improve their health—and save on high costs that would be required during a hospital visit.
To enter the contest, you'll need to register as a competitor at kaggle.com, and then log into the competition site at the Heritage Health Prize site. The first two tables of data will be released later today, with additional tables released on May 4 and June 4. Progress prizes will be awarded to the leading teams on August 13, February 13 (2012) and September 4 (2014) and announced at the following Strata conferences. The Grand Prize will be awarded to the leading team as of April 13, 2013 provided their predictive model exceeds the Accuracy Threshold (to be announced on May 4); otherwise a $500,000 consolation prize will be awarded instead. Full competition rules can be found at the Heritage Health competition site, linked below.
This is a great opportunity for any data scientist to test their mettle, and if the Netflix prize is any indicator, I'm sure there will be intense competion as the deadlines near for each of the four prizes.
Heritage Health Prize: Description
Comments