Why Multitasking Doesn’t Work — And What Actually Works - LifeHack

Deep reinforcement learning-based radiotherapy machine parameter optimization with a multitask policy network