Summary
Project Summary and Abstract Health care decisions are made in challenging clinical settings where patients and physicians have to con- tend with complex trade-offs across a range of longitudinal and time-to-event quality-of-life outcomes and mortal- ity. The recent advent of Machine Learning has strongly contributed to a substantial interest in the development of novel risk prediction tools. Most proposed risk scores, however, are limited by a univariate perspective on a single outcome. The scores remain agnostic to other outcomes co-occurring with the outcome of interest or precluding it as
