Hindsight Credit Assignment

Authors:
  • TMThomas Mesnard
    Thomas Mesnard
  • MAMohammad Azar
    Mohammad Azar
  • BPBilal Piot
    Bilal Piot
  • NHNicolas Heess
    Nicolas Heess
  • HHHado van Hasselt
    Hado van Hasselt
  • GWGreg Wayne
    Greg Wayne
  • SSSatinder Singh
    Satinder Singh
  • DPDoina Precup
    Doina Precup
  • RMRemi Munos
    Remi Munos

Abstract
We consider the problem of efficient credit assignment in reinforcement learning. In order to efficiently and meaningfully utilize new data, we propose to explicitly assign credit to past decisions based on the likelihood of them having led to the observed outcome. This approach uses new information in hindsight, rather than employing foresight. Somewhat surprisingly, we show that value functions can be rewritten through this lens, yielding a new family of algorithms. We study the properties of these algorithms, and empirically show that they successfully address important credit assignment challenges, through a set of illustrative tasks.
All Related Versions

This publication has 2 references indexed in Scilit: