This is a Kaggle competition.  The goal of this competition is to predict if a person has any of three medical conditions. You are being asked to predict if the person has one or more of any of the three medical conditions (Class 1), or none of the three medical conditions (Class 0). 

It is thus at first sight a binary classification problem.  However, the task is to provide probabilities, not just a binary prediction.  Probabilistic classification is described in Introduction to Probabilistic Classification: A Machine Learning Perspective.  

Results also should be calibrated, as various classification approaches introduce biases.  A paper from Cornell University researchers Alexandru Niculescu-Mizil and Rich Caruana, Predicting Good Probabilities With Supervised Learning, discusses these biases and proposes the best approaches to correct them.  

I will try to test the results from that paper, and see if the recommended approaches do improve predicted probabailities.  I certainly don't expect to win any competitions, but I hope to learn a lot about probabilistic classification and calibration.  We'll see.

The competition is done!  See my summary on my blog.  Spoiler: I did not do well in the competition, but I did learn a lot.