I was a researcher in the machine learning group for the Cardiovascular Research Grid project. We worked with the cardiologists who are custodians of a dataset involving patients with ICD implants. Our objective was to come up with robust, interpretable predictors that predicted either high risk of ICD utilization , or sudden cardiac death. An important challenge was to create predictors that both utilized both survival data and did not become biased due to not-missing-at-random data.