Simple Statistical Gradient-Following Algorithms for Connectionist Reinforcement Learning. Williams


  • The REINFORCE paper
  • Does not explicitly compute gradients
  • Deals mostly with immediate reward (so not sequential decision making it seems), deals with delayed reward in a restricted case
  • Results framed in terms of connectionist networks (ANNs), with no hidden layer
    • But can be generalized as long as they are differentiable
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: