Dernoncourt Franck, Nemati Shamim, Kassis Elias Baedorf, Ghassemi Mohammad Mahdi
Massachusetts Institute of Technology, Cambridge, MA, USA
Beth Israel Medical Center, Boston, MA, USA
Algorithms and features in medical studies contain many “knobs” that govern the learning process from a high-level perspective: they are called hyperparameters, and investigators typically tune them by hand. In this case study, we present three mathematically grounded techniques to automatically optimize hyperparameters, and demonstrate their use in the problem of outcome prediction for ICU patients who suffer from sepsis.
医学研究中的算法和特征包含许多从高层次角度控制学习过程的“旋钮”：它们被称为超参数，研究人员通常手动调整这些参数。在本案例研究中，我们提出了三种基于数学的技术来自动优化超参数，并展示它们在脓毒症重症监护病房（ICU）患者预后预测问题中的应用。
