PropertyValue
nif:beginIndex
  • 0 (xsd:integer)
nif:broaderContext
nif:endIndex
  • 198 (xsd:integer)
nif:isString
  • During the training time, as in [7], the exploration rate (proportion of actions chosen randomly) decreases from an initial 1.0 to 0.05 in the first million time steps and stays fixed at that value.
rdf:type