Skip to content
voiage

voiage.methods.belief_state_information.belief_state_information_value

belief_state_information_value

Section titled “belief_state_information_value”
belief_state_information_value([positional or keyword] specification: Mapping[str, object] = None) -> BeliefStateInformationResult

Evaluate an exact finite control-transition-observe-update problem.

The adaptive Bellman policy is compared with a matched no-information policy. The result reports action-dependent-learning and dual-control diagnostics, but deliberately does not manufacture a unique additive dual-control value.

Parameters:

  • specification Mapping[str, object]

Returns: BeliefStateInformationResult