3 ms·Improve exploration in reinforcement learning, acts as a regularizer in supervised learning.by tonybeltramelli 10y agoImprove exploration in reinforcement learning, acts as a regularizer in supervised learning.