Home > marl > agent > actfuns > teamgreedyact.m

teamgreedyact

PURPOSE ^

Implements deterministic greedy action selection for full-state, joint action Q-learners

SYNOPSIS ^

function [a, action] = teamgreedyact(a, state, actions, rewards, params)

DESCRIPTION ^

Implements deterministic greedy action selection for full-state, joint action Q-learners
  [A, ACTION] = TEAMGREEDYACT(A, STATE, ACTIONS, REWARDS, PARAMS)
  Implements team-greedy action selection. Supports discrete states and
  actions, with 1 action variable per agent.

  Values on the agent learning parameters:
   usesocconv      - whether to use social conventions to break the ties
       in maximization. Default 1 (use social conventions). The ordering
       is agents from n to 1, actions from 1 to max.


  See also agent_act

CROSS-REFERENCE INFORMATION ^

This function calls: This function is called by:
Generated on Wed 04-Aug-2010 16:55:08 by m2html © 2005