Increasing Entropy to Boost Policy Gradient Performance on Personalization Tasks.

Andrew Starnes, Anton Dereventsov, Clayton G. Webster

Browse the full ICDM paper archive.