1 paper · 1 filter
Donghwan Lee
Periodic target updates in Q-learning and soft target updates in actor-critic methods are empirically well established stabilization mechanisms, but their precise theoretical expla…